JP3903928B2 - Audio switching device - Google Patents

Audio switching device Download PDF

Info

Publication number
JP3903928B2
JP3903928B2 JP2003048260A JP2003048260A JP3903928B2 JP 3903928 B2 JP3903928 B2 JP 3903928B2 JP 2003048260 A JP2003048260 A JP 2003048260A JP 2003048260 A JP2003048260 A JP 2003048260A JP 3903928 B2 JP3903928 B2 JP 3903928B2
Authority
JP
Japan
Prior art keywords
transmission
reception
noise power
signal
voice
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
JP2003048260A
Other languages
Japanese (ja)
Other versions
JP2004260491A (en
Inventor
実 福島
敏 杉本
博昭 竹山
靖久 井平
恵一 ▲吉▼田
彰洋 菊池
章 寺澤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Panasonic Electric Works Co Ltd
Original Assignee
Matsushita Electric Works Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Matsushita Electric Works Ltd filed Critical Matsushita Electric Works Ltd
Priority to JP2003048260A priority Critical patent/JP3903928B2/en
Publication of JP2004260491A publication Critical patent/JP2004260491A/en
Application granted granted Critical
Publication of JP3903928B2 publication Critical patent/JP3903928B2/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Landscapes

  • Interconnected Communication Systems, Intercoms, And Interphones (AREA)
  • Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
  • Telephone Function (AREA)

Description

【０００１】
【発明の属する技術分野】
本発明は、集合住宅のインターホンシステムなどに用いられる拡声通話端末において使用される音声切換装置に関するものである。
【０００２】
【従来の技術】
従来より、通話時にハンドセットを持つ必要がなく、通話端末から離れた通話者に対して相手側の通話端末から伝送されてくる音声信号をスピーカにより送出し、かつ、上記通話者の発する音声をマイクロホンにより集音して相手側通話端末へ伝送することで半二重通話を可能とする拡声通話システムが提供されている。このような拡声通話システムにおいては、その構成要素であるスピーカ−マイクロホン間の音響結合や、音声信号の伝送路が２線の形態で構成される場合に必要となる２線−４線変換ハイブリッド回路におけるインピーダンスの不整合により生じる送話信号路から受話信号路への回り込み、及び相手側の通話端末におけるスピーカ−マイクロホン間の音響結合等によって通話路上に閉ループが形成され、この閉ループの一巡利得が１倍以上になるとハウリングが生じ、ハウリングが生じた場合には通話を継続することができないため、これを抑圧する手段が必要となる。
【０００３】
そこで従来は、送話信号及び受話信号を監視することにより通話状態が受話状態または送話状態の何れであるかを判別し、判別された通話状態に応じて送話信号路又は受話信号路の少なくとも一方に損失を挿入することにより、閉ループの一巡利得を低減させてハウリングを防止する音声切換装置（いわゆる音声スイッチ）が拡声通話端末に広く用いられてきた。音声切換装置の基本的な動作は、送話信号及び受話信号のパワーを推定し、これらの大小関係を比較して瞬時パワーの小さい側に対して所定量の損失を挿入するというものである。
【０００４】
図２２は特許文献１に開示されている従来の音声切換装置を示すブロック図である。この従来例は、拡声通話端末のマイクロホン（図示せず）で集音する送話信号を回線へ伝送するための送話側信号経路Ｌ１に損失を挿入する送話側損失挿入手段１と、回線から受信した受話信号を拡声通話端末のスピーカ（図示せず）へ伝送するための受話側信号経路Ｌ２に損失を挿入する受話側損失挿入手段２と、送話側損失挿入手段１に入力される送話信号を取り出して増幅する送話偏重モード設定用増幅器６と、受話側損失挿入手段２に入力される受話信号を取り出して増幅する受話偏重モード設定用増幅器７と、送話偏重モード設定用増幅器６並びに受話偏重モード設定用増幅器７で増幅された送話信号及び受話信号に基づいて通話モードを推定するとともに該推定結果に応じて送話側損失挿入手段１並びに受話側損失挿入手段２が各経路Ｌ１，Ｌ２に挿入する損失量を制御して通話モードを送話モード、受話モード並びに中立モードに切り換える挿入損失量制御手段３と、送話信号に含まれる近端側の雑音パワーを推定する近端側雑音パワー推定手段４と、受話信号に含まれる遠端側の雑音パワーを推定する遠端側雑音パワー推定手段５と、遠端側雑音パワー並びに近端側雑音パワーの各推定値ＰＦｎ，ＰＮｎに応じて送話偏重モード設定用増幅器６並びに受話偏重モード設定用増幅器７の各利得を調整する偏重モード制御手段８’とを備えている。
【０００５】
遠端側雑音パワー推定手段５並びに近端側雑音パワー推定手段４は、何れも立ち上がりが緩やかであり且つ立ち下がりが急峻な特性をもつ積分回路又はデジタルフィルタ等によって実現され、遠端側雑音パワー推定手段５では受話信号中に定常的に存在する暗騒音（背景雑音）パワーを推定し、近端側雑音パワー推定手段４では送話信号中に定常的に存在する雑音パワーを推定する。
【０００６】
偏重モード制御手段８’は、遠端側雑音パワーの推定値ＰＦｎが近端側雑音パワーの推定値ＰＮｎよりも充分に大きい値であれば（ＰＦｎ≫ＰＮｎ）、送話偏重モード設定用増幅器６の利得ＧＴをＧ（＞０）［ｄＢ］、受話偏重モード設定用増幅器７の利得ＧＲを０［ｄＢ］とすることで通話モードを送話偏重モードに設定し、近端側雑音パワーの推定値ＰＮｎが遠端側雑音パワーの推定値ＰＦｎよりも充分に大きい値であれば（ＰＮｎ≫ＰＦｎ）、受話偏重モード設定用増幅器７の利得ＧＲをＧ［ｄＢ］、送話偏重モード設定用増幅器６の利得ＧＴを０［ｄＢ］とすることで通話モードを受話偏重モードに設定し、遠端側雑音パワーの推定値ＰＦｎと近端側雑音パワーの推定値ＰＮｎの差が充分に大きい値でなければ、受話偏重モード設定用増幅器７並びに送話偏重モード設定用増幅器６の各利得ＧＲ，ＧＴを０［ｄＢ］として中立モードに設定する。
【０００７】
すなわち、遠端側の周囲騒音レベルと近端側の周囲騒音レベルとの差が大きい場合、送話信号及び受話信号を監視して通話状態を推定する挿入損失量制御手段３では、例えば遠端側の周囲騒音レベルが大きい状況においては常に受話状態と判定し、近端側の周囲騒音レベルが大きい状況においては常に送話状態と判定してしまい、実際の通話状態に関係なく、受話状態又は送話状態の何れか一方に通話状態を固定してしまう現象（所謂片倒れ）が生じることがある。
【０００８】
これに対して上記従来例では、上述のように偏重モード制御手段８’が遠端側雑音パワーの推定値ＰＦｎと近端側雑音パワーの推定値ＰＮｎを比較し、遠端側雑音パワーの推定値ＰＦｎの方が充分に大きい場合は挿入損失量制御手段３で監視する送話信号を送話偏重モード設定用増幅器６で利得Ｇ［ｄＢ］だけ増幅することにより、挿入損失量制御手段３が送話状態と判定し易い状態（送話偏重モード）に設定し、反対に近端側雑音パワーの推定値ＰＮｎの方が充分に大きい場合は挿入損失量制御手段３で監視する受話信号を受話偏重モード設定用増幅器７で利得Ｇ［ｄＢ］だけ増幅することにより、挿入損失量制御手段３が受話状態と判定し易い状態（受話偏重モード）に設定することにより、上記片倒れを抑制して良好な切換特性を得ることができるようになっている。
【０００９】
【特許文献１】
特開２００２−３５９５８０号公報（段落００８１−段落００８５、第２１図）
【００１０】
【発明が解決しようとする課題】
ところが上記従来例においては、例えば近端側から一方的に発声している場合に近端側雑音パワーの推定値ＰＮｎが徐々に増加し、近端側及び遠端側における各周囲騒音レベルの差が小さくても偏重モード制御手段８’により受話偏重モードに設定され、近端側の話者が発声しているにもかかわらず挿入損失量制御手段３が送話モードから受話モードに誤って切り換えてしまい、音声が途切れたり不自然な抑揚が生じる虞がある。同様に、遠端側から一方的に発声している場合には遠端側雑音パワーの推定値ＰＦｎが徐々に増加して偏重モード制御手段８’により送話偏重モードに設定され、遠端側の話者が発声しているにもかかわらず挿入損失量制御手段３が受話モードから送話モードに誤って切り換えてしまい、音声が途切れたり不自然な抑揚が生じる虞がある。
【００１１】
本発明は上記事情に鑑みて為されたものであり、その目的は、誤った通話モードに設定されるのを防いで音声の途切れや抑揚を抑えることができる音声切換装置を提供することにある。
【００１２】
【課題を解決するための手段】
請求項１の発明は、上記目的を達成するために、マイクロホン及びスピーカを有する拡声通話端末が他の通話端末又は拡声通話端末に有線で接続される拡声通話系の前記拡声通話端末に用いられ、前記マイクロホンで集音する送話信号を回線へ伝送するための送話側信号経路に損失を挿入する送話側損失挿入手段と、回線から受信した受話信号を前記スピーカへ伝送するための受話側信号経路に損失を挿入する受話側損失挿入手段と、前記送話側損失挿入手段に入力される送話信号を取り出して増幅する送話偏重モード設定用増幅手段と、前記受話側損失挿入手段に入力される受話信号を取り出して増幅する受話偏重モード設定用増幅手段と、前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段で増幅された送話信号及び受話信号に基づいて通話モードを推定するとともに該推定結果に応じて前記送話側損失挿入手段並びに受話側損失挿入手段が前記経路に挿入する損失量を制御して前記通話モードを送話モードと受話モードに切り換える挿入損失量制御手段と、送話信号に含まれる近端側の雑音パワーを推定する近端側雑音パワー推定手段と、受話信号に含まれる遠端側の雑音パワーを推定する遠端側雑音パワー推定手段と、遠端側雑音パワー並びに近端側雑音パワーの各推定値に応じて前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の各利得を調整する偏重モード制御手段とを備え、前記挿入損失量制御手段が、前記送話側損失挿入手段への入力信号の瞬時パワーを推定する第１の瞬時パワー推定部と、前記受話側損失挿入手段への入力信号の瞬時パワーを推定する第２の瞬時パワー推定部と、前記送話側損失挿入手段への入力点から前記送話側損失挿入手段並びに回線側での回り込みを経て前記受話側損失挿入手段への入力点へ帰還する系の利得に応じて決定される値を係数にもつ回線帰還利得乗算部と、前記受話側損失挿入手段への入力点から前記受話側損失挿入手段並びに音響側での回り込みを経て前記送話側損失挿入手段への入力点へ到る経路の利得に応じて決定される値を係数にもつ音響帰還利得乗算部と、第２の瞬時パワー推定部の出力信号を音響帰還利得乗算部へ入力して得られる出力信号と第１の瞬時パワー推定部の出力信号との大小関係を比較する第１の比較器と、第１の瞬時パワー推定部の出力信号を回線帰還利得乗算部へ入力して得られる出力信号と第２の瞬時パワー推定部の出力信号との大小関係を比較する第２の比較器と、送話信号の音声区間を検出する送話信号音声区間検出部と、受話信号の音声区間を検出する受話信号音声区間検出部と、第１の比較器及び第２の比較器の比較結果と送話信号音声区間検出部及び受話信号音声区間検出部の検出結果とに基づいて通話状態を判定し送話側損失挿入手段及び受話側損失挿入手段の挿入損失量を制御する挿入損失量分配処理部とを具備し、前記偏重モード制御手段は、遠端側並びに近端側の各雑音パワーの推定値の大小関係と前記送話信号音声区間検出部の検出結果に基づいて、前記送話偏重モード設定用増幅手段及び前記受話偏重モード設定用増幅手段の各利得をほぼ等しくする中立モードと前記受話偏重モード設定用増幅手段の利得を前記送話偏重モード設定用増幅手段の利得よりも増大させる受話偏重モードの何れかに設定することを特徴とする。
【００１３】
請求項２の発明は、請求項１の発明において、前記偏重モード制御手段は、前記近端側雑音パワーの推定値が前記遠端側雑音パワーの推定値と所定の第１の雑音パワー比係数との積以上の値となり且つ前記送話信号音声区間検出部によって音声区間が検出されない状態の継続時間が第１の所定時間以上となったときに受話偏重モード設定用増幅手段の利得を送話偏重モード設定用増幅手段の利得よりも増大させて受話偏重モードに設定することを特徴とする。
【００１４】
請求項３の発明は、上記目的を達成するために、マイクロホン及びスピーカを有する拡声通話端末が他の通話端末又は拡声通話端末に有線で接続される拡声通話系の前記拡声通話端末に用いられ、前記マイクロホンで集音する送話信号を回線へ伝送するための送話側信号経路に損失を挿入する送話側損失挿入手段と、回線から受信した受話信号を前記スピーカへ伝送するための受話側信号経路に損失を挿入する受話側損失挿入手段と、前記送話側損失挿入手段に入力される送話信号を取り出して増幅する送話偏重モード設定用増幅手段と、前記受話側損失挿入手段に入力される受話信号を取り出して増幅する受話偏重モード設定用増幅手段と、前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段で増幅された送話信号及び受話信号に基づいて通話モードを推定するとともに該推定結果に応じて前記送話側損失挿入手段並びに受話側損失挿入手段が前記経路に挿入する損失量を制御して前記通話モードを送話モードと受話モードに切り換える挿入損失量制御手段と、送話信号に含まれる近端側の雑音パワーを推定する近端側雑音パワー推定手段と、受話信号に含まれる遠端側の雑音パワーを推定する遠端側雑音パワー推定手段と、遠端側雑音パワー並びに近端側雑音パワーの各推定値に応じて前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の各利得を調整する偏重モード制御手段とを備え、前記挿入損失量制御手段が、前記送話側損失挿入手段への入力信号の瞬時パワーを推定する第１の瞬時パワー推定部と、前記受話側損失挿入手段への入力信号の瞬時パワーを推定する第２の瞬時パワー推定部と、前記送話側損失挿入手段への入力点から前記送話側損失挿入手段並びに回線側での回り込みを経て前記受話側損失挿入手段への入力点へ帰還する系の利得に応じて決定される値を係数にもつ回線帰還利得乗算部と、前記受話側損失挿入手段への入力点から前記受話側損失挿入手段並びに音響側での回り込みを経て前記送話側損失挿入手段への入力点へ到る経路の利得に応じて決定される値を係数にもつ音響帰還利得乗算部と、第２の瞬時パワー推定部の出力信号を音響帰還利得乗算部へ入力して得られる出力信号と第１の瞬時パワー推定部の出力信号との大小関係を比較する第１の比較器と、第１の瞬時パワー推定部の出力信号を回線帰還利得乗算部へ入力して得られる出力信号と第２の瞬時パワー推定部の出力信号との大小関係を比較する第２の比較器と、送話信号の音声区間を検出する送話信号音声区間検出部と、受話信号の音声区間を検出する受話信号音声区間検出部と、第１の比較器及び第２の比較器の比較結果と送話信号音声区間検出部及び受話信号音声区間検出部の検出結果とに基づいて通話状態を判定し送話側損失挿入手段及び受話側損失挿入手段の挿入損失量を制御する挿入損失量分配処理部とを具備し、前記偏重モード制御手段は、遠端側並びに近端側の各雑音パワーの推定値の大小関係と前記受話信号音声区間検出部の検出結果に基づいて、前記送話偏重モード設定用増幅手段及び前記受話偏重モード設定用増幅手段の各利得をほぼ等しくする中立モードと前記送話偏重モード設定用増幅手段の利得を前記受話偏重モード設定用増幅手段の利得よりも増大させる送話偏重モードの何れかに設定することを特徴とする。
【００１５】
請求項４の発明は、請求項３の発明において、前記偏重モード制御手段は、前記遠端側雑音パワーの推定値が前記近端側雑音パワーの推定値と所定の第２の雑音パワー比係数との積以上の値となり且つ前記受話信号音声区間検出部によって音声区間が検出されない状態の継続時間が第２の所定時間以上となったときに送話偏重モード設定用増幅手段の利得を受話偏重モード設定用増幅手段の利得よりも増大させて送話偏重モードに設定することを特徴とする。
【００１６】
請求項５の発明は、請求項１〜４の何れかの発明において、前記送話側又は受話側の音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記しきい値並びに前記所定値を外部から設定可能としたことを特徴とする。
【００１７】
請求項６の発明は、請求項１〜４の何れかの発明において、前記送話側又は受話側の音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記瞬時パワー推定部は立ち上がりが急峻で立ち下がりが緩やかな特性を有するフィルタからなり、前記雑音パワー推定部は立ち上がりが緩やかで立ち下がりが急峻な特性を有するフィルタからなり、該フィルタの前記特性を決定するパラメータを外部から設定可能としたことを特徴とする。
【００１８】
請求項７の発明は、請求項５又は６の発明において、前記しきい値及び所定値又は前記パラメータを前記送話側又は受話側の音声区間検出部に対して外部から個別に設定可能としたことを特徴とする。
【００１９】
請求項８の発明は、請求項５又は６の発明において、前記しきい値及び所定値又は前記パラメータを前記他の通話端末の種類に応じて各々異なる値に設定可能としたことを特徴とする。
【００２０】
請求項９の発明は、請求項１〜８の何れかの発明において、前記送話側又は受話側の音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記近端側雑音パワー推定手段を前記送話側音声区間検出部を構成する前記雑音パワー推定部で兼用するとともに、前記遠端側雑音パワー推定手段を前記受話側音声区間検出部を構成する前記雑音パワー推定部で兼用したことを特徴とする。
【００２１】
請求項１０の発明は、請求項２の発明において、前記偏重モード制御手段は、前記継続時間が前記第１の所定時間未満であっても、前記近端側雑音パワーの推定値が前記遠端側雑音パワーの推定値と前記第１の雑音パワー比係数よりも大きい所定の第３の雑音パワー比係数との積以上となる状態が前記第１の所定時間よりも長い所定の第３の所定時間以上継続したときには受話偏重モードに設定することを特徴とする。
【００２２】
請求項１１の発明は、請求項４の発明において、前記偏重モード制御手段は、前記継続時間が前記第２の所定時間未満であっても、前記遠端側雑音パワーの推定値が前記近端側雑音パワーの推定値と前記第２の雑音パワー比係数よりも大きい所定の第４の雑音パワー比係数との積以上となる状態が前記第２の所定時間よりも長い所定の第４の所定時間以上継続したときには送話偏重モードに設定することを特徴とする。
【００２３】
請求項１２の発明は、請求項３の発明において、前記受話側損失挿入手段の出力点から近端側の音響エコー経路を介して前記送話側損失挿入手段の入力点へ帰還する経路の音響側帰還利得を推定する音響側帰還利得推定手段を備え、前記偏重モード制御手段は、音響側帰還利得の推定値が所定の条件を満たさなければ送話偏重モードに移行しないことを特徴とする。
【００２４】
請求項１３の発明は、請求項１２の発明において、前記偏重モード制御手段は、前記遠端側雑音パワーの推定値が前記近端側雑音パワーの推定値と前記第２の雑音パワー比係数との積以上であり、前記受話信号音声区間検出部の検出結果が非音声区間であり、且つ前記音響側帰還利得の推定値が所定のしきい値未満である状態が所定時間以上継続したときに送話偏重モードに設定することを特徴とする。
【００２５】
請求項１４の発明は、請求項１の発明において、前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たさなければ受話偏重モードに移行しないことを特徴とする。
【００２６】
請求項１５の発明は、請求項１４の発明において、前記偏重モード制御手段は、前記近端側雑音パワーの推定値が前記遠端側雑音パワーの推定値と前記第１の雑音パワー比係数との積以上であり、前記送話信号音声区間検出部の検出結果が非音声区間であり、且つ前記回線側帰還利得の推定値が所定のしきい値未満である状態が所定時間以上継続したときに受話偏重モードに設定することを特徴とする。
【００２７】
請求項１６の発明は、請求項１２の発明において、前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たさなければ受話偏重モードに移行せず、前記送話側及び受話側の各音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記送話側及び受話側の各音声区間検出部は、前記音響側並びに回線側の各帰還利得推定手段の推定値に応じて前記しきい値及び所定値が変更されることを特徴とする。
【００２８】
請求項１７の発明は、請求項１２の発明において、前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たす場合には受話偏重モードに設定し、前記送話側及び受話側の各音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記瞬時パワー推定部は立ち上がりが急峻で立ち下がりが緩やかな特性を有するフィルタからなり、前記雑音パワー推定部は立ち上がりが緩やかで立ち下がりが急峻な特性を有するフィルタからなり、前記送話側及び受話側の各音声区間検出部は、前記音響側並びに回線側の各帰還利得推定手段の推定値に応じて前記特性を決定するパラメータが変更されることを特徴とする。
【００２９】
請求項１８の発明は、請求項１７の発明において、前記しきい値及び所定値と前記パラメータを前記送話側並びに受話側の各音声区間検出部に対して外部から個別に設定可能としたことを特徴とする。
【００３０】
請求項１９の発明は、請求項１７の発明において、前記しきい値及び所定値と前記パラメータを前記他の通話端末の種類に応じて各々異なる値に設定可能としたことを特徴とする。
【００３１】
請求項２０の発明は、請求項１２の発明において、前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たさなければ受話偏重モードに移行せず、前記近端側雑音パワーの推定値と前記遠端側雑音パワーの推定値との差と、前記音響側並びに回線側の帰還利得推定値を参照して前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の利得を適応的に更新することを特徴とする。
【００３２】
請求項２１の発明は、請求項２０の発明において、前記偏重モード制御手段により適応的に更新される前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の利得に上限値を設定したことを特徴とする。
【００３３】
【発明の実施の形態】
（実施形態１）
本実施形態は、基本的な構成が従来例と共通であるから共通の構成要素には同一の符号を付して説明を省略する。
【００３４】
本実施形態では、図１に示すように送話側損失挿入手段１への入力信号の瞬時パワーを推定する第１の瞬時パワー推定部３１と、受話側損失挿入手段２への入力信号の瞬時パワーを推定する第２の瞬時パワー推定部３２と、送話側損失挿入手段１への入力点から送話側損失挿入手段１並びに回線側での回り込みを経て受話側損失挿入手段２への入力点へ帰還する系の利得に応じて決定される値を係数にもつ回線帰還利得乗算部３３と、受話側損失挿入手段２への入力点から受話側損失挿入手段２並びに音響側での回り込みを経て送話側損失挿入手段１への入力点へ到る経路の利得に応じて決定される値を係数にもつ音響帰還利得乗算部３４と、第２の瞬時パワー推定部３２の出力信号を音響帰還利得乗算部３４へ入力して得られる出力信号と第１の瞬時パワー推定部３１の出力信号との大小関係を比較する第１の比較器３５と、第１の瞬時パワー推定部３１の出力信号を回線帰還利得乗算部３３へ入力して得られる出力信号と第２の瞬時パワー推定部３２の出力信号との大小関係を比較する第２の比較器３６と、送話信号の音声区間を検出する送話信号音声区間検出部３７と、受話信号の音声区間を検出する受話信号音声区間検出部３８と、第１の比較器３５及び第２の比較器３６の比較結果と送話信号音声区間検出部３７及び受話信号音声区間検出部３８の検出結果とに基づいて通話状態を判定し送話側損失挿入手段１及び受話側損失挿入手段２の挿入損失量を制御する挿入損失量分配処理部３０とを挿入損失量制御手段３が具備し、偏重モード制御手段８は、遠端側並びに近端側の各雑音パワーの推定値の大小関係と送話信号音声区間検出部３７の検出結果に基づいて、送話偏重モード設定用増幅器６及び受話偏重モード設定用増幅器７の各利得ＧＴ，ＧＲをほぼ等しくする（例えば、０［ｄＢ］とする）中立モードと受話偏重モード設定用増幅器７の利得ＧＲを送話偏重モード設定用増幅器６の利得ＧＴよりも増大させる受話偏重モードの何れかに設定する点に特徴がある。
【００３５】
第１の比較器３５では、第１の瞬時パワー推定部３１からの出力信号と第２の瞬時パワー推定部３２からの出力信号（第１の瞬時パワー推定値）を音響帰還利得乗算手段３４へ入力して得られる出力信号とを比較しており、瞬時パワーの推定値が音響帰還利得乗算手段３４の出力信号以上の場合に出力信号Ｃ１が１となり、瞬時パワーの推定値が音響帰還利得乗算手段３４の出力信号未満の場合に出力信号Ｃ１が０となる。また、第２の比較器３６では、第１の瞬時パワー推定部３１の出力信号を回線帰還利得乗算手段３３へ入力して得られる出力信号と第２の瞬時パワー推定部３２の出力信号（第２の瞬時パワー推定値）とを比較しており、回線側帰還利得乗算手段３３の出力信号が第２の瞬時パワー推定値以上の場合に出力信号Ｃ２が１となり、回線側帰還利得乗算手段３３の出力信号が第２の瞬時パワー推定値未満の場合に出力信号Ｃ２が０となる。
【００３６】
第１及び第２の瞬時パワー推定部３１，３２は、立ち上がりが急峻であり且つ立ち下がりが緩やかな特性をもつ積分回路又はデジタルフィルタによって実現され、それぞれ送話側損失挿入手段１への入力信号及び受話側損失挿入手段２への入力信号をそれぞれ送話偏重モード設定用増幅器６並びに受話偏重モード設定用増幅器７で増幅した信号の瞬時パワーを推定するものである。
【００３７】
図２は送話信号音声区間検出部３７及び受話信号音声区間検出部３８の具体構成を示すブロック図である。送話信号音声区間検出部３７は、送話信号（受話信号音声区間検出部３８においては受話信号、以下かっこ内は受話信号音声区間検出部３８の場合を表す）を参照して近端側（遠端側）における背景雑音レベルを推定する背景雑音パワー推定部３７ａ（３８ａ）と、第１の瞬時パワー推定部３１（３２）で推定される瞬時パワー推定値Ｐｓ並びに背景雑音パワー推定部３７ａ（３８ａ）で推定される背景雑音パワー推定値Ｐｎに基づいて送話側又は受話側の損失挿入手段１，２への入力信号（以下、これらを総称して「参照信号」と呼ぶ）ｉｎが音声信号であるか非音声信号であるかを判定し、音声信号と判定した場合には判定結果（判定フラグ）ＳＤＦ１（ＳＤＦ２）を１とし、非音声信号と判定した場合には判定結果ＳＤＦ１（ＳＤＦ２）を０とするとともに、判定結果ＳＤＦ１（ＳＤＦ２）が更新されるまで前回の判定結果ＳＤＦ１（ＳＤＦ２）を保持する音声／非音声判定部３７ｂ（３８ｂ）とを具備する。なお、背景雑音パワー推定部３７ａ（３８ａ）は、立ち上がりが緩やかであり、且つ立ち下がりが急峻な特性をもつ積分回路又はデジタルフィルタによって構成され、参照信号ｉｎを参照して逐次背景雑音パワー推定値Ｐｎを更新し、更新するまでは前の推定値Ｐｎを保持している。
【００３８】
一方音声／非音声判定部３７ｂ（３８ｂ）は、例えば、第１の瞬時パワー推定部３１（３２）から出力される瞬時パワー推定値Ｐｓを所定のしきい値Ｐs0と比較し、瞬時パワー推定値Ｐｓと背景雑音パワー推定部３７ａ（３８ａ）から出力される背景雑音パワー推定値Ｐｎとの比Ｐｓ／Ｐｎを所定のしきい値δと比較するとともに、瞬時パワー推定値Ｐｓがしきい値Ｐs0よりも大きく（Ｐｓ＞Ｐs0）、且つ前記比Ｐｓ／Ｐｎがしきい値δよりも大きい（Ｐｓ／Ｐｎ＞δ）場合に音声信号と判定して判定結果ＳＤＦ１（ＳＤＦ２）を１とし、その他の場合に非音声信号と判定して判定結果ＳＤＦ１（ＳＤＦ２）を０とする。ここで、しきい値Ｐs0は音声信号の最小レベルを規定するしきい値であり、しきい値δは音声信号レベルと背景雑音レベルとの最小比を規定するしきい値である。
【００３９】
本実施形態における挿入損失量分配処理部３０では、第１及び第２の比較器３５，３６の比較結果Ｃ１，Ｃ２と送話信号音声区間検出部３７及び受話信号音声区間検出部３８の検出結果ＳＤＦ１，ＳＤＦ２を参照して通話状態を判定し、送話側損失挿入手段１及び受話側損失挿入手段２の挿入損失量を決定する。
【００４０】
次に、図３のフローチャートを参照して本実施形態における偏重モード制御手段８の動作を説明する。
【００４１】
まず、送話信号音声区間検出部３７の検出結果（判定結果）を判断し（ステップ１）、音声区間が検出されていなければ（ＳＤＦ１＝０であれば）、中立モードから受話偏重モードへの移行判定を行い（ステップ２）、移行条件をクリアしているか否かを判断する（ステップ３）。なお、ここでの移行条件は従来例における中立モードから受話偏重モードへの移行条件と同一である。そして、上記移行条件をクリアしていれば受話偏重モード設定用増幅器７の利得ＧＲをＧ［ｄＢ］、送話偏重モード設定用増幅器６の利得ＧＴを０［ｄＢ］とすることで受話偏重モードに設定する（ステップ６）。
【００４２】
一方、ステップ１において送話信号音声区間検出部３７により音声区間が検出されている場合、若しくはステップ３において移行条件をクリアしていない場合には、中立モードから送話偏重モードへの移行判定を行い（ステップ４）、移行条件をクリアしているか否かを判断する（ステップ５）。なお、ここでの移行条件も従来例における中立モードから送話偏重モードへの移行条件と同一である。そして、上記移行条件をクリアしていれば送話偏重モード設定用増幅器６の利得ＧＴをＧ［ｄＢ］、受話偏重モード設定用増幅器７の利得ＧＲを０［ｄＢ］とすることで送話偏重モードに設定し（ステップ７）、移行条件をクリアしなければ、送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲを何れも０［ｄＢ］とすることで中立モードに設定する（ステップ８）。
【００４３】
而して本実施形態においては、近端側端末（本実施形態の音声切換装置を具備している拡声通話端末）の話者が一方的に発声している場合に、送話信号音声区間検出部３７で音声区間が検出されているときには２つの雑音パワー推定手段４，５の推定値ＰＦｎ，ＰＮｎにかかわらず、偏重モード制御手段８が受話偏重モードに設定しないから、挿入損失量制御手段３により送話側損失挿入手段１による挿入損失量が最大且つ受話側損失挿入手段２による挿入損失量が最小となるように制御されて通話モードが送話モードに設定されるため、近端側の話者が発声しているにもかかわらず挿入損失量制御手段３が中間モードから受話モードに誤って切り換えてしまうのを防ぎ、音声が途切れたり不自然な抑揚が生じるのを抑えることができる。なお、中間モードとは、送話側及び受話側の損失挿入手段１，２の各挿入損失量が略同一である状態であって、いわゆるアイドルモード（例えば、送話側及び受話側とも無音の状況）として扱われる場合である。
【００４４】
（実施形態２）
本実施形態は偏重モード制御手段８に特徴があり、これ以外の構成については実施形態１と同一であるから図示並びに説明を省略する。
【００４５】
本実施形態における偏重モード制御手段８は、図４に示すように遠端側雑音パワーの推定値ＰＦｎと所定の第１の雑音パワー比係数Ｘ１との積を求める乗算器８２と、この積と近端側雑音パワーの推定値ＰＮｎとを比較する第１のコンパレータ８１と、送話信号音声区間検出部３７の検出結果ＳＤＦ１を反転するインバータ８３と、第１のコンパレータ８１の比較結果とインバータ８３で反転された検出結果ＳＤＦ１の論理積ｉを求めるアンドゲート８４と、この論理積ｉが１のときにインクリメントされるとともに０のときにリセットされるカウンタからなる計時手段８５と、計時手段８５によって計時される継続時間（カウント値）Ｔ’と第１の所定時間Ｔ１を比較する第２のコンパレータ８６と、第２のコンパレータ８６の比較結果Ｚに応じて送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲをそれぞれＧ［ｄＢ］又は０［ｄＢ］に設定する利得設定部８７とを具備する。
【００４６】
偏重モード制御手段８の動作を具体的に説明すると、遠端側雑音パワーの推定値ＰＦｎと第１の雑音パワー比係数Ｘ１の積と近端側雑音パワーの推定値ＰＮｎとが第１のコンパレータ８１において比較され、ＰＦｎ×Ｘ１＜ＰＮｎのときに１、ＰＦｎ×Ｘ１≧ＰＮｎのときに０の比較結果がアンドゲート８４に出力される。そして、近端側背景雑音パワーの推定値ＰＮｎが前記積よりも大きい値であり、且つ送話信号音声区間検出部３７により音声区間が検出されていない場合にのみアンドゲート８４の出力（論理積）ｉが１となり、計時手段８５における継続時間Ｔ’がインクリメントされ、近端側背景雑音パワーの推定値ＰＮｎが前記積以下になるか、あるいは送話信号音声区間検出部３７により音声区間が検出されるか、何れかの条件が成立するとアンドゲート８４の出力ｉが０となって計時手段８５における継続時間Ｔ’がリセットされる。計時手段８５による継続時間Ｔ’は第２のコンパレータ８６にて第１の所定時間Ｔ１と比較されており、継続時間Ｔ’が第１の所定時間Ｔ１を超えたときに第２のコンパレータ８６の出力（比較結果）Ｚが１となり、継続時間Ｔ’が第１の所定時間Ｔ１を超えなければ出力Ｚは０となる。そして、第２のコンパレータ８６の出力Ｚが１の場合に利得設定部８７が送話偏重モード設定用増幅器６の利得ＧＴを０［ｄＢ］、受話偏重モード設定用増幅器７の利得ＧＲをＧ［ｄＢ］に設定することで受話偏重モードに設定され、出力Ｚが０の場合に利得設定部８７が送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲを何れも０［ｄＢ］に設定することで中立モードに設定される。
【００４７】
このように本実施形態の偏重モード制御手段８では、近端側雑音パワーの推定値ＰＮｎが遠端側雑音パワーの推定値ＰＦｎと第１の雑音パワー比係数Ｘ１との積よりも大きい値となり且つ送話信号音声区間検出部３７によって音声区間が検出されない状態の継続時間Ｔ’を計時し、継続時間Ｔ’が第１の所定時間Ｔ１以上となったときに受話偏重モード設定用増幅手段７の利得ＧＲをＧ［ｄＢ］、送話偏重モード設定用増幅手段６の利得ＧＴを０［ｄＢ］として受話偏重モードに設定するため、第１の雑音パワー比係数Ｘ１と第１の所定時間Ｔ１を適切な値に設定することで偏重モード制御手段８による中立モードから受話偏重モードへの移行のしやすさを調整することができ、本実施形態の音声切換装置を搭載する拡声通話端末の周囲騒音に対する切換特性を任意に設定することができる。
【００４８】
（実施形態３）
本実施形態は、図５に示すように基本的な構成が従来例及び実施形態１と共通であるから共通の構成要素には同一の符号を付して説明を省略し、本実施形態の特徴となる構成についてのみ説明する。本実施形態では、偏重モード制御手段８が、遠端側並びに近端側の各雑音パワーの推定値ＰＦｎ，ＰＮｎの大小関係と受話信号音声区間検出部３８の検出結果ＳＤＦ２に基づいて送話偏重モード又は中立モードの何れかに設定する点に特徴がある。
【００４９】
次に、図６のフローチャートを参照して本実施形態における偏重モード制御手段８の動作を説明する。
【００５０】
まず、受話信号音声区間検出部３８の検出結果（判定結果）を判断し（ステップ１）、音声区間が検出されていなければ（ＳＤＦ２＝０であれば）、中立モードから送話偏重モードへの移行判定を行い（ステップ２）、移行条件をクリアしているか否かを判断する（ステップ３）。なお、ここでの移行条件は従来例における中立モードから送話偏重モードへの移行条件と同一である。そして、上記移行条件をクリアしていれば受話偏重モード設定用増幅器７の利得ＧＲを０［ｄＢ］、送話偏重モード設定用増幅器６の利得ＧＴをＧ［ｄＢ］とすることで送話偏重モードに設定する（ステップ６）。
【００５１】
一方、ステップ１において受話信号音声区間検出部３８により音声区間が検出されている場合、若しくはステップ３において移行条件をクリアしていない場合には、中立モードから受話偏重モードへの移行判定を行い（ステップ４）、移行条件をクリアしているか否かを判断する（ステップ５）。なお、ここでの移行条件も従来例における中立モードから受話偏重モードへの移行条件と同一である。そして、上記移行条件をクリアしていれば送話偏重モード設定用増幅器６の利得ＧＴを０［ｄＢ］、受話偏重モード設定用増幅器７の利得ＧＲをＧ［ｄＢ］とすることで受話偏重モードに設定し（ステップ７）、移行条件をクリアしなければ、送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲを何れも０［ｄＢ］とすることで中立モードに設定する（ステップ８）。
【００５２】
而して本実施形態においては、遠端側端末（本実施形態の音声切換装置を具備している拡声通話端末と拡声通話システムを構成する他の通話端末）の話者が一方的に発声している場合に、受話信号音声区間検出部３８で音声区間が検出されているときには２つの雑音パワー推定手段４，５の推定値ＰＦｎ，ＰＮｎにかかわらず、偏重モード制御手段８が送話偏重モードに設定しないから、挿入損失量制御手段３により送話側損失挿入手段１による挿入損失量が最小且つ受話側損失挿入手段２による挿入損失量が最大となるように制御されて通話モードが受話モードに設定されるため、遠端側の話者が発声しているにもかかわらず挿入損失量制御手段３が中立モードから送話モードに誤って切り換えてしまうのを防ぎ、音声が途切れたり不自然な抑揚が生じるのを抑えることができる。
【００５３】
なお、実施形態１と同様に、偏重モード制御手段８が、遠端側並びに近端側の各雑音パワーの推定値ＰＦｎ，ＰＮｎの大小関係と送話信号音声区間検出部３７の検出結果ＳＤＦ１に基づいて受話偏重モード又は中立モードの何れかに設定するようにすれば、挿入損失量制御手段３が中立モードから送話モード又は受話モードに誤って切り換えてしまうのを防いで音声が途切れたり不自然な抑揚が生じるのをより確実に抑えることができる。
【００５４】
（実施形態４）
本実施形態は偏重モード制御手段８に特徴があり、これ以外の構成については実施形態３と同一であるから図示並びに説明を省略する。
【００５５】
本実施形態における偏重モード制御手段８は、図７に示すように近端側雑音パワーの推定値ＰＮｎと所定の第２の雑音パワー比係数Ｘ２との積を求める乗算器８２’と、この積と遠端側雑音パワーの推定値ＰＦｎとを比較する第３のコンパレータ８１’と、受話信号音声区間検出部３８の検出結果ＳＤＦ２を反転するインバータ８３’と、第３のコンパレータ８１’の比較結果とインバータ８３’で反転された検出結果ＳＤＦ２の論理積ｉ’を求めるアンドゲート８４’と、この論理積ｉ’が１のときにインクリメントされるとともに０のときにリセットされるカウンタからなる計時手段８５’と、計時手段８５’による計時時間（カウント値）Ｔ”と第２の所定時間Ｔ２を比較する第４のコンパレータ８６’と、第４のコンパレータ８６’の比較結果Ｚ’に応じて送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲをそれぞれＧ［ｄＢ］又は０［ｄＢ］に設定する利得設定部８７’とを具備する。
【００５６】
偏重モード制御手段８の動作を具体的に説明すると、近端側雑音パワーの推定値ＰＮｎと第２の雑音パワー比係数Ｘ２の積と遠端側雑音パワーの推定値ＰＦｎとが第３のコンパレータ８１’において比較され、ＰＮｎ×Ｘ２＜ＰＦｎのときに１、ＰＮｎ×Ｘ２≧ＰＦｎのときに０の比較結果がアンドゲート８４’に出力される。そして、遠端側背景雑音パワーの推定値ＰＦｎが前記積よりも大きい値であり、且つ受話信号音声区間検出部３８により音声区間が検出されていない場合にのみアンドゲート８４’の出力（論理積）ｉ’が１となり、計時手段８５’における継続時間Ｔ”がインクリメントされ、遠端側背景雑音パワーの推定値ＰＦｎが前記積よりも大きくなくなるか、あるいは受話信号音声区間検出部３８により音声区間が検出されるか、何れかの条件が成立するとアンドゲート８４’の出力ｉ’が０となって計時手段８５’における継続時間Ｔ”がリセットされる。計時手段８５’による継続時間Ｔ”は第４のコンパレータ８６’にて第２の所定時間Ｔ２と比較されており、継続時間Ｔ”が第２の所定時間Ｔ２を超えたときに第４のコンパレータ８６’の出力（比較結果）Ｚ’が１となり、継続時間Ｔ”が第２の所定時間Ｔ２を超えなければ出力Ｚ’は０となる。そして、第４のコンパレータ８６’の出力Ｚ’が１の場合に利得設定部８７’が送話偏重モード設定用増幅器６の利得ＧＴをＧ［ｄＢ］、受話偏重モード設定用増幅器７の利得ＧＲを０［ｄＢ］に設定することで送話偏重モードに設定され、出力Ｚ’が０の場合に利得設定部８７’が送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲを何れも０［ｄＢ］に設定することで中立モードに設定される。
【００５７】
このように本実施形態の偏重モード制御手段８では、遠端側雑音パワーの推定値ＰＦｎが近端側雑音パワーの推定値ＰＮｎと第２の雑音パワー比係数Ｘ２との積よりも大きい値となり且つ受話信号音声区間検出部３８によって音声区間が検出されない状態の継続時間Ｔ”を計時し、継続時間Ｔ”が第２の所定時間Ｔ２以上となったときに受話偏重モード設定用増幅手段７の利得ＧＲを０［ｄＢ］、送話偏重モード設定用増幅手段６の利得ＧＴをＧ［ｄＢ］として送話偏重モードに設定するため、第２の雑音パワー比係数Ｘ２と第２の所定時間Ｔ２を適切な値に設定することで偏重モード制御手段８による中立モードから送話偏重モードへの移行のしやすさを調整することができ、本実施形態の音声切換装置を搭載する拡声通話端末の周囲騒音に対する切換特性を任意に設定することができる。
【００５８】
（実施形態５）
ところで、上述の実施形態１〜４の音声切換装置は個々の手段をハードウェアで構成することは勿論可能ではあるが、デジタル・シグナル・プロセッサ（ＤＳＰ）のような単一のプロセッサを用い、ＤＳＰのハードウェアを専用のソフトウェアで制御することによって上記各手段を実現することが望ましい。そして、本実施形態の音声切換装置ＶＳはＤＳＰと専用のソフトウェアの組み合わせで実現されるものである。
【００５９】
図８は本実施形態の音声切換装置ＶＳを備えた拡声通話端末の概略構成を示している。ＤＳＰ１００は、一連の通話処理機能と、報知音や警報音あるいは警報音声を生成する機能を有している。音声切換装置ＶＳはＤＳＰ１００の機能の一部として構成される。また、図中の２００はＣＰＵであって、例えば相手側の通話端末からの呼出を検出してＤＳＰ１００に通話処理を開始させる制御を行う機能をＲＯＭ２０１に格納された専用のソフトウェアをメモリに読み込んで実行することによって実現している。なお、ＤＳＰ１００とＣＰＵ２００とはシリアルポートあるいはパラレルポートなどの通信用のインタフェースを介して接続されている。
【００６０】
本実施形態における拡声通話端末では、通話を開始する際にＣＰＵ２００からＤＳＰ１００に対して、送話信号音声区間検出部３７又は受話信号音声区間検出部３８において音声区間の検出に用いるしきい値Ｐs0，δのデータをＣＰＵ２００から前記通信用インタフェースを介してＤＳＰ１００に送信し、これをＤＳＰ１００に設けたしきい値設定手段１０１により音声区間検出部３７，３８を実現するソフトウェアモジュールに対して初期設定する。これらしきい値Ｐs0，δの値（データ）が初期化された後、ＤＳＰ１００が音声切換装置ＶＳやエコーキャンセラ等を起動して通話処理を開始し、ＣＰＵ２００が拡声通話端末のハードウェアを制御して相手側の通話端末との間の通話路を形成する。なお、しきい値設定手段１０１もソフトウェアで実現されるものである。
【００６１】
本実施形態は上述のように構成したものであって、送話信号音声区間検出部３７又は受話信号音声区間検出部３８において音声区間の検出に用いるしきい値Ｐs0，δを外部（ＣＰＵ２００）から設定可能としたので、音声切換装置ＶＳの汎用性が高まり、種々の拡声通話端末に容易に適応可能になる。
【００６２】
（実施形態６）
図９は本実施形態の音声切換装置ＶＳを備えた拡声通話端末の概略構成を示しており、実施形態５と共通の構成要素には同一の符号を付している。
【００６３】
本実施形態における拡声通話端末では、通話を開始する際にＣＰＵ２００からＤＳＰ１００に対して、送話信号音声区間検出部３７又は受話信号音声区間検出部３８において音声区間の検出に用いるパラメータ、具体的には瞬時パワー推定部９１や背景雑音パワー推定部９２の演算で使用される時定数（デジタルフィルタの立ち上がり及び立ち下がりの特性を決定するパラメータ）のデータをＣＰＵ２００から通信用インタフェースを介してＤＳＰ１００に送信し、これをＤＳＰ１００に設けた時定数設定手段１０２により音声区間検出部３７，３８を実現するソフトウェアモジュールに対して初期設定する。これらの時定数が初期化された後、ＤＳＰ１００が音声切換装置ＶＳやエコーキャンセラ等を起動して通話処理を開始し、ＣＰＵ２００が拡声通話端末のハードウェアを制御して相手側の通話端末との間の通話を形成する。なお、時定数設定手段１０２もＤＳＰ１００のハードウェアをソフトウェアで制御することによって実現されるものである。
【００６４】
本実施形態は上述のように構成したものであって、送話信号音声区間検出部３７又は受話信号音声区間検出部３８において音声区間の検出に用いる時定数を外部（ＣＰＵ２００）から設定可能としたので、音声切換装置ＶＳの汎用性が高まり、種々の拡声通話端末に容易に適応可能になる。
【００６５】
（実施形態７）
図１０は本実施形態の音声切換装置ＶＳを備えた拡声通話端末の概略構成を示しており、実施形態５と共通の構成要素には同一の符号を付している。
【００６６】
本実施形態における拡声通話端末では、送話信号音声区間検出部３７に対するしきい値Ｐs0，δ（これらを「送話側しきい値」と呼ぶ）並びに時定数（これを「送話側時定数」と呼ぶ）と、受話信号音声区間検出部３８に対するしきい値Ｐs0，δ（これらを「受話側しきい値」と呼ぶ）並びに時定数（これを「受話側時定数」と呼ぶ）とをフラッシュメモリ１１０に格納している。
【００６７】
一方ＤＳＰ１００は、通信用インタフェースを介してＣＰＵ２００からコマンドを受信するとともに該コマンドを解釈して必要な処理を行うコマンド処理手段１０３と、音声切換装置ＶＳに対して送話側及び受話側の各しきい値と時定数を音声区間検出部３７，３８を実現するソフトウェアモジュールに対して設定するパラメータ設定手段１０４とを備えている。但し、コマンド処理手段１０３やパラメータ設定手段１０４もソフトウェアで実現可能である。
【００６８】
通話を開始する際には、まずＣＰＵ２００からＤＳＰ１００に対して通話開始要求コマンドを通信用インタフェースを介して送信し、この通話開始要求コマンドを受け取ったコマンド処理手段１０３がその内容を解釈し、送話信号音声区間検出部３７並びに受話信号音声区間検出部３８において用いる上記パラメータの格納場所を示すアドレスデータをパラメータ設定手段１０４に与える。パラメータ設定手段１０４では受け取ったアドレスデータに基づいて送話側及び受話側のしきい値並びに時定数のデータをフラッシュメモリ１１０から読み込み、音声区間検出部３７，３８を実現するソフトウェアモジュールに対して初期設定する。これら初期設定が完了した後、パラメータ設定手段１０４はパラメータの初期化が完了したことを示すフラグをセットし、このフラグのセットを受けてコマンド処理手段１０３が応答コマンドをＣＰＵ２００に対して送信する。
【００６９】
而して、近端側端末（本実施形態における拡声通話端末）のマイクロホン（図示せず）から音声切換装置ＶＳにおける送話信号の参照点までの伝達特性と遠端側端末のマイクロホン（図示せず）から受信信号の参照点までの伝達特性は一般に異なるが、本実施形態においては送話信号音声区間検出部３７並びに受話信号送信音声区間検出部１０に対するしきい値Ｐs0，δ及び時定数が外部（ＣＰＵ２００）から各々個別に設定可能であるから、上述のような２つの伝達特性の差を簡単に補正することができる。
【００７０】
（実施形態８）
図１１は本実施形態の音声切換装置ＶＳを備えた拡声通話端末の概略構成を示しており、実施形態７と共通の構成要素には同一の符号を付している。
【００７１】
本発明の音声切換装置が用いられる通話システムとして、マンションなどの集合住宅における玄関ロビーに設置されたロビーインターホン、並びに各住戸の玄関に設置されるドアホンと、各住戸内に設置される親機との間で相互に拡声通話を行うものがある。このような通話システムにおいては、実施形態７で説明した受話信号の参照点までの伝達特性が相手側の端末によって異なっている。すなわち、音声切換装置ＶＳを備えた拡声通話端末が上記親機である場合にドアホンとの通話における受話信号の伝達特性と、ロビーインターホンとの通話における受話信号の伝達特性とは通話の線路長等の影響で異なっており、そのために送話信号の伝達特性と受話信号の伝達特性との差も当然に相手側の通話端末毎に異なることになる。
【００７２】
そこで本実施形態では、相手側の通話端末の種類（上述の例であればドアホンとロビーインターホン）に応じて各々異なる値に設定可能としており、具体的にはフラッシュメモリ１１０の２つの領域Ｍ１，Ｍ２にそれぞれドアホン用のパラメータ（送話側及び受話側のしきい値並びに時定数）とロビーインターホン用のパラメータを格納している。そして、ＣＰＵ２００が呼出相手の通話端末の種類に応じたパラメータのアドレスをコマンドで指定し、パラメータ設定手段１０４が受け取ったアドレスデータに基づいて相手側の通話端末の種類に応じた送話側及び受話側のしきい値並びに時定数のデータをフラッシュメモリ１１０の領域Ｍ１，Ｍ２から選択して読み込み、音声区間検出部３７，３８を実現するソフトウェアモジュールに対して初期設定する。
【００７３】
而して、本実施形態においては送話信号音声区間検出部３７並びに受話信号音声区間検出部３８に対するしきい値Ｐs0，δ及び時定数が相手側の通話端末毎に外部（ＣＰＵ２００）から各々個別に設定可能であるから、相手側の通話端末毎に異なる２つの伝達特性の差を簡単に補正することができる。
【００７４】
（実施形態９）
本実施形態のブロック図を図１２に示す。本実施形態が実施形態１及び実施形態３と異なる点は、近端側雑音パワー推定手段４を送話側音声区間検出部３７を構成する背景雑音パワー推定部３７ａ（図２参照）で兼用するとともに、遠端側雑音パワー推定手段５を受話側音声区間検出部３８を構成する背景雑音パワー推定部３８ａ（図２参照）で兼用している点にある。なお、これ以外の構成及び動作については実施形態１又は実施形態３と同一であるから説明は省略する。
【００７５】
すなわち、本実施形態では実施形態１あるいは実施形態３に比較して構成を簡略化することができ、またＤＳＰ１００とソフトウェアで構成する場合にあってはソフトウェアにおける演算量を低減することができる。
【００７６】
（実施形態１０）
本実施形態は偏重モード制御手段８に特徴があり、これ以外の構成については実施形態１と同一であるから図示並びに説明を省略する。
【００７７】
本実施形態における偏重モード制御手段８は基本的な構成が実施形態２と共通であって、図１３に示すように第１のコンパレータ８１、乗算器８２、インバータ８３、アンドゲート８４、計時手段８５、第２のコンパレータ８６並びに利得設定部８７に加えて、遠端側雑音パワーの推定値ＰＦｎと所定の第３の雑音パワー比係数Ｘ３との積を求める乗算器１８２と、この積と近端側雑音パワーの推定値ＰＮｎとを比較する第５のコンパレータ１８１と、第５のコンパレータ１８１の出力が１のときにインクリメントされるとともに０のときにリセットされるカウンタからなる計時手段１８３と、計時手段１８３による計時時間（カウント値）Ｔ２’と第３の所定時間Ｔ３を比較する第６のコンパレータ１８４と、第２のコンパレータ８６の比較結果Ｙ１と第６のコンパレータ１８４の比較結果Ｙ２の論理和Ｚを求めるオアゲート８８とを具備する。なお、第３の雑音パワー比係数Ｘ３は第１の雑音パワー比係数Ｘ１の数倍以上の値に設定され、第３の所定時間Ｔ３は第１の所定時間Ｔ１よりも大きい値に設定される。
【００７８】
次に本実施形態における偏重モード制御手段８の動作を説明する。但し、実施形態２と同一構成の部分についての動作説明は省略し、実施形態２に対して本実施形態で追加された構成の動作及び全体の動作についてのみ説明する。
【００７９】
遠端側雑音パワーの推定値ＰＦｎと第３の雑音パワー比係数Ｘ３の積と近端側雑音パワーの推定値ＰＮｎとが第５のコンパレータ１８１において比較され、ＰＦｎ×Ｘ３＜ＰＮｎのときに１、ＰＦｎ×Ｘ３≧ＰＮｎのときに０の比較結果が計時手段１８３に出力される。そして、近端側背景雑音パワーの推定値ＰＮｎが前記積よりも大きい値のときには、計時手段１８３における継続時間Ｔ２’がインクリメントされ、近端側背景雑音パワーの推定値ＰＮｎが前記積よりも大きくなくなるか、あるいは送話信号音声区間検出部３７により音声区間が検出されるか、何れかの条件が成立すると計時手段１８３における継続時間Ｔ２’がリセットされる。計時手段１８３による継続時間Ｔ２’は第６のコンパレータ１８４にて第３の所定時間Ｔ３と比較されており、継続時間Ｔ２’が第３の所定時間Ｔ３を超えたときに第６のコンパレータ１８４の出力（比較結果）Ｙ２が１となり、継続時間Ｔ２’が第３の所定時間Ｔ３を超えなければ出力Ｙ２は０となる。
【００８０】
そして、第６のコンパレータ１８４の出力Ｙ２と第２のコンパレータ８６の出力Ｙ１がオアゲート８８に入力されており、少なくとも何れか一方の出力Ｙ１，Ｙ２が１であればオアゲート８８の出力Ｚが１となり、両出力Ｙ１，Ｙ２が０であればオアゲート８８の出力Ｚも０となる。さらに、オアゲート８８の出力Ｚが１の場合に利得設定部８７が送話偏重モード設定用増幅器６の利得ＧＴを０［ｄＢ］、受話偏重モード設定用増幅器７の利得ＧＲをＧ［ｄＢ］に設定することで受話偏重モードに設定され、出力Ｚが０の場合に利得設定部８７が送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲを何れも０［ｄＢ］に設定することで中立モードに設定される。
【００８１】
而して、実施形態２においては、近端側の通話端末の周囲に高レベルの非定常騒音（例えば、テレビやラジオの放送音など）が存在する場合に送話信号音声区間検出部３７が音声区間を誤検出することで受話偏重モードに設定され難くなり、通話モードが送話モードに固定されてしまう現象（所謂片倒れ）が発生する虞がある。これに対して本実施形態における偏重モード制御手段８は、継続時間Ｔ’が第１の所定時間Ｔ１未満であっても、近端側雑音パワーの推定値ＰＮｎが遠端側雑音パワーの推定値ＰＦｎと第３の雑音パワー比係数Ｘ３との積よりも大きい状態が第３の所定時間Ｔ３以上継続したときに受話偏重モードに設定するので、上述のように近端側の通話端末の周囲に高レベルの非定常騒音が存在する場合でも受話偏重モードに設定され難くなるのを防いで片倒れの発生が防止できる。
【００８２】
（実施形態１１）
本実施形態は偏重モード制御手段８に特徴があり、これ以外の構成については実施形態３と同一であるから図示並びに説明を省略する。
【００８３】
本実施形態における偏重モード制御手段８は基本的な構成が実施形態４と共通であって、図１４に示すように第３のコンパレータ１８１’、乗算器８２’、インバータ８３’、アンドゲート８４’、計時手段８５’、第４のコンパレータ８６’並びに利得設定部８７’に加えて、近端側雑音パワーの推定値ＰＮｎと所定の第４の雑音パワー比係数Ｘ４との積を求める乗算器１８２’と、この積と遠端側雑音パワーの推定値ＰＦｎとを比較する第７のコンパレータ１８１’と、第７のコンパレータ１８１’の出力が１のときにインクリメントされるとともに０のときにリセットされるカウンタからなる計時手段１８３’と、計時手段１８３’による計時時間（カウント値）Ｔ２”と第４の所定時間Ｔ４を比較する第８のコンパレータ１８４’と、第４のコンパレータ８６’の比較結果Ｙ３と第８のコンパレータ１８４’の比較結果Ｙ４の論理和Ｚ’を求めるオアゲート８８’とを具備する。なお、第４の雑音パワー比係数Ｘ４は第２の雑音パワー比係数Ｘ２の数倍以上の値に設定され、第４の所定時間Ｔ４は第２の所定時間Ｔ２よりも大きい値に設定される。
【００８４】
次に本実施形態における偏重モード制御手段８の動作を説明する。但し、実施形態４と同一構成の部分についての動作説明は省略し、実施形態４に対して本実施形態で追加された構成の動作及び全体の動作についてのみ説明する。
【００８５】
近端側雑音パワーの推定値ＰＮｎと第４の雑音パワー比係数Ｘ４の積と遠端側雑音パワーの推定値ＰＦｎとが第７のコンパレータ１８１’において比較され、ＰＮｎ×Ｘ４＜ＰＦｎのときに１、ＰＮｎ×Ｘ４≧ＰＦｎのときに０の比較結果が計時手段１８３’に出力される。そして、遠端側背景雑音パワーの推定値ＰＦｎが前記積よりも大きい値のときに、計時手段１８３’における継続時間Ｔ２”がインクリメントされ、遠端側背景雑音パワーの推定値ＰＦｎが前記積よりも大きくなくなるか、あるいは受話信号音声区間検出部３８により音声区間が検出されるか、何れかの条件が成立すると計時手段１８３’における継続時間Ｔ２”がリセットされる。計時手段１８３’による継続時間Ｔ２”は第８のコンパレータ１８４’にて第４の所定時間Ｔ４と比較されており、継続時間Ｔ２”が第４の所定時間Ｔ４を超えたときに第８のコンパレータ１８４’の出力（比較結果）Ｙ４が１となり、継続時間Ｔ２”が第４の所定時間Ｔ４を超えなければ出力Ｙ４は０となる。
【００８６】
そして、第８のコンパレータ１８４’の出力Ｙ４と第４のコンパレータ８６’の出力Ｙ３がオアゲート８８’に入力されており、少なくとも何れか一方の出力Ｙ３，Ｙ４が１であればオアゲート８８’の出力Ｚ’が１となり、両出力Ｙ３，Ｙ４が０であればオアゲート８８’の出力Ｚ’も０となる。さらに、オアゲート８８’の出力Ｚ’が１の場合に利得設定部８７’が送話偏重モード設定用増幅器６の利得ＧＴをＧ［ｄＢ］、受話偏重モード設定用増幅器７の利得ＧＲを０［ｄＢ］に設定することで送話偏重モードに設定され、出力Ｚ’が０の場合に利得設定部８７’が送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲを何れも０［ｄＢ］に設定することで中立モードに設定される。
【００８７】
而して、実施形態４においては、遠端側の通話端末の周囲に高レベルの非定常騒音（例えば、風きり音や工事に伴う騒音など）が存在する場合に受話信号音声区間検出部３８が音声区間を誤検出することで送話偏重モードに設定され難くなり、通話モードが受話モードに固定されてしまう現象（所謂片倒れ）が発生する虞がある。これに対して本実施形態における偏重モード制御手段８は、継続時間Ｔ”が第２の所定時間Ｔ２未満であっても、遠端側雑音パワーの推定値ＰＦｎが近端側雑音パワーの推定値ＰＮｎと第４の雑音パワー比係数Ｘ４との積よりも大きい状態が第４の所定時間Ｔ４以上継続したときには送話偏重モードに移行するので、上述のように遠端側の通話端末の周囲に高レベルの非定常騒音が存在する場合でも送話偏重モードに確実に移行するから、片倒れの発生が防止できる。
【００８８】
（実施形態１２）
本実施形態の音声切換装置は、図１５に示すように実施形態９の構成に加えて、受話側損失挿入手段２の出力点から近端側の音響エコー経路ＨＡＣを介して送話側損失挿入手段１の入力点へ帰還する経路の音響側帰還利得αを推定する音響側帰還利得推定手段１１を備え、音響側帰還利得αの推定値｜α’｜が所定の条件を満たさなければ偏重モード制御手段８が受話偏重モードに移行しない点に特徴がある。
【００８９】
音響側帰還利得推定手段１１では、送話側損失挿入手段１の入力信号（送話信号）の短時間における時間平均パワーを推定するとともに、受話側損失挿入手段２の入力信号（受話信号）の短時間における時間平均パワーを推定し、さらに音響側帰還経路ＨＡＣにて想定される最大遅延時間において受話側損失挿入手段２の出力信号の時間平均パワーの推定値の最小値を求め、この最小値で送話側損失挿入手段１の入力信号の時間平均パワーの推定値を除算した値を音響側帰還利得αの推定値｜α’｜としている。
【００９０】
上述の各実施形態においては、送話偏重モードのときには送話偏重モード設定用増幅器６の利得を増大させているために所謂受話ブロッキングが生じ易くなる。ここで受話ブロッキングとは、近端側が無音の状態で遠端側より音声が入力されたときに、近端側のスピーカ−マイクロホン間の音響結合によって生じる音響エコー信号により第１及び第２の比較器３５，３６の何れか若しくは両方の比較結果が受話状態から送話状態へ反転することにより、挿入損失量分配処理部３０において推定される通話モードが受話モードではなくなり、受話側損失挿入手段２に損失が挿入されるため、遠端側から入力された音声を近端側で受聴する際に不自然な途切れを生じたり、損失量が大きい場合には全く聞き取れなくなる現象を言う。これに対して本実施形態では、音響側帰還利得αの推定値｜α’｜が所定のしきい値α０以下となる状態が一定時間以上継続しなければ偏重モード制御手段８が送話偏重モードに移行しないため、受話ブロッキングを防止できる。
【００９１】
なお、音響側帰還利得αの推定値｜α’｜の求め方は、上記方法に限定されるものではなく、本出願人が既に出願した特許出願の明細書に記載した従来既知の方法を採用しても構わない。
【００９２】
（実施形態１３）
本実施形態は偏重モード制御手段８に特徴があり、これ以外の構成については実施形態１２と同一であるから図示並びに説明を省略する。
【００９３】
本実施形態においては、偏重モード制御手段８が、遠端側雑音パワーの推定値ＰＦｎが近端側雑音パワーの推定値ＰＮｎと第２の雑音パワー比係数Ｘ２との積以上であり、受話信号音声区間検出手段１０の検出結果ＳＤＦ２が音声区間でなく、且つ音響側帰還利得αの推定値｜α’｜が所定のしきい値α０以下である状態が所定時間以上継続しなければ送話偏重モードに移行しないことにより、受話ブロッキングの発生を抑えている点に特徴がある。
【００９４】
本実施形態における偏重モード制御手段８は基本的な構成が実施形態４と共通であって、図１６に示すように第３のコンパレータ８１’、乗算器８２’、インバータ８３’、アンドゲート８４’、計時手段８５’、第４のコンパレータ８６’並びに利得設定部８７’に加えて、音響側帰還利得αの推定値｜α’｜をしきい値α０と比較する第１０のコンパレータ１８５’と、アンドゲート８４’の出力と第１０のコンパレータ１８５’の比較結果の論理積を求める第２のアンドゲート１８６’とを具備し、第２のアンドゲート１８６’の出力を計時手段８５’の入力としている。
【００９５】
次に本実施形態における偏重モード制御手段８の動作を説明する。但し、実施形態４と同一構成の部分についての動作説明は省略し、実施形態４に対して本実施形態で追加された構成の動作及び全体の動作についてのみ説明する。
【００９６】
音響側帰還利得αの推定値｜α’｜としきい値α０とが第１０のコンパレータ１８５’で比較され、推定値｜α’｜がしきい値α０未満のときに１、推定値｜α’｜がしきい値α０以上のときに０の比較結果が第２のアンドゲート１８６’に出力される。したがって、遠端側背景雑音パワーの推定値ＰＦｎが前記積よりも大きい値であり、受話信号音声区間検出部３８により音声区間が検出されておらず、且つ音響側帰還利得αの推定値｜α’｜がしきい値α０未満の場合にのみ計時手段８５’における継続時間Ｔ”がインクリメントされ、遠端側背景雑音パワーの推定値ＰＦｎが前記積よりも大きくなくなるか、受話信号音声区間検出部３８により音声区間が検出されるか、あるいは音響側帰還利得αの推定値｜α’｜がしきい値α０以上となるかの何れかの条件が成立すると計時手段８５’における継続時間Ｔ”がリセットされる。そして、計時手段８５’による継続時間Ｔ”が第２の所定時間Ｔ２を超えたときに第４のコンパレータ８６’の出力Ｚ’が１となって利得設定部８７’が送話偏重モード設定用増幅器６の利得ＧＴをＧ［ｄＢ］、受話偏重モード設定用増幅器７の利得ＧＲを０［ｄＢ］に設定することで送話偏重モードに設定され、継続時間Ｔ”が第２の所定時間Ｔ２を超えなければ出力Ｚ’は０となって利得設定部８７’が送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲを何れも０［ｄＢ］に設定することで中立モードに設定される。
【００９７】
而して、音響側帰還利得αの推定値｜α’｜がしきい値α０以上であるときには他の条件にかかわらず偏重モード制御手段８が常に中立モードに設定し、音響側帰還利得αが相対的に大きい状況では送話偏重モードに設定しないため、受話ブロッキングの発生を抑えることができる。さらに、しきい値α０を設定可変なパラメータとして利用できるので、機器の設置環境に応じて機器の性能を様々に設定することが可能である。
【００９８】
（実施形態１４）
本実施形態の音声切換装置は、図１７に示すように実施形態９の構成に加えて、送話側損失挿入手段１の出力点から遠端側の回線エコー経路ＨＬＩＮを介して受話側損失挿入手段２の入力点へ帰還する経路の回線側帰還利得βを推定する回線側帰還利得推定手段１２を備え、回線側帰還利得βの推定値｜β’｜が所定の条件を満たさなければ偏重モード制御手段８が送話偏重モードに移行しない点に特徴がある。
【００９９】
回線側帰還利得推定手段１２では、受話側損失挿入手段２の入力信号（受話信号）の短時間における時間平均パワーを推定するとともに、送話側損失挿入手段１の入力信号（送話信号）の短時間における時間平均パワーを推定し、さらに回線側帰還経路ＨＬＩＮにて想定される最大遅延時間において送話側損失挿入手段１の出力信号の時間平均パワーの推定値の最小値を求め、この最小値で受話側損失挿入手段２の入力信号の時間平均パワーの推定値を除算した値を回線側帰還利得βの推定値｜β’｜としている。
【０１００】
上述の各実施形態においては、受話偏重モードのときには受話偏重モード設定用増幅器７の利得を増大させているために所謂送話ブロッキングが生じ易くなる。ここで送話ブロッキングとは、遠端側が無音の状態で近端側より音声が入力されたときに、遠端側における音響結合又は２線−４線変換回路における信号の回り込みによって生じる回線エコー信号により第１及び第２の比較器３５，３６の何れか若しくは両方の比較結果が送話状態から受話状態へ反転することにより、挿入損失量分配処理部３０において推定される通話モードが送話モードではなくなり、送話側損失挿入手段１に損失が挿入されるため、近端側から入力された音声を遠端側で受聴する際に不自然な途切れを生じたり、損失量が大きい場合には全く聞き取れなくなる現象を言う。これに対して本実施形態では、回線側帰還利得βの推定値｜β’｜が所定値β０以下となる状態が一定時間以上継続しなければ偏重モード制御手段８が送話偏重モードに移行しないため、送話ブロッキングを防止できる。
【０１０１】
なお、回線側帰還利得βの推定値｜β’｜の求め方は、上記方法に限定されるものではなく、本出願人が既に出願した特許出願の明細書に記載した従来既知の方法を採用しても構わない。
【０１０２】
（実施形態１５）
本実施形態は偏重モード制御手段８に特徴があり、これ以外の構成については実施形態１３と同一であるから図示並びに説明を省略する。
【０１０３】
本実施形態においては、偏重モード制御手段８が、近端側雑音パワーの推定値ＰＮｎが遠端側雑音パワーの推定値ＰＦｎと第１の雑音パワー比係数Ｘ１との積以上であり、送話信号音声区間検出部３７の検出結果ＳＤＦ１が音声区間でなく、且つ回線側帰還利得βの推定値｜β’｜が所定のしきい値β０以下である状態が所定時間以上継続したときに受話偏重モードに設定することにより、送話ブロッキングの発生を抑えている点に特徴がある。
【０１０４】
本実施形態における偏重モード制御手段８は基本的な構成が実施形態２と共通であって、図１８に示すように第１のコンパレータ８１、乗算器８２、インバータ８３、アンドゲート８４、計時手段８５、第２のコンパレータ８６並びに利得設定部８７に加えて、回線側帰還利得βの推定値｜β’｜をしきい値β０と比較する第９のコンパレータ１８５と、アンドゲート８４の出力と第９のコンパレータ１８５の比較結果の論理積を求める第２のアンドゲート１８６とを具備し、第２のアンドゲート１８６の出力を計時手段８５の入力としている。
【０１０５】
次に本実施形態における偏重モード制御手段８の動作を説明する。但し、実施形態２と同一構成の部分についての動作説明は省略し、実施形態２に対して本実施形態で追加された構成の動作及び全体の動作についてのみ説明する。
【０１０６】
回線側帰還利得βの推定値｜β’｜としきい値β０とが第９のコンパレータ１８５で比較され、推定値｜β’｜がしきい値β０未満のときに１、推定値｜β’｜がしきい値β０以上のときに０の比較結果が第２のアンドゲート１８６に出力される。したがって、近端側背景雑音パワーの推定値ＰＮｎが前記積よりも大きい値であり、送話信号音声区間検出部３７により音声区間が検出されておらず、且つ回線側帰還利得βの推定値｜β’｜がしきい値β０未満の場合にのみ計時手段８５における継続時間Ｔ’がインクリメントされ、近端側背景雑音パワーの推定値ＰＮｎが前記積以下になるか、送話信号音声区間検出部３７により音声区間が検出されるか、あるいは回線側帰還利得βの推定値｜β’｜がしきい値β０以上となるかの何れかの条件が成立すると計時手段８５における継続時間Ｔ’がリセットされる。そして、計時手段８５による継続時間Ｔ’が第１の所定時間Ｔ１を超えたときに第２のコンパレータ８６の出力Ｚが１となって利得設定部８７が送話偏重モード設定用増幅器６の利得ＧＴを０［ｄＢ］、受話偏重モード設定用増幅器７の利得ＧＲをＧ［ｄＢ］に設定することで受話偏重モードに設定され、継続時間Ｔ’が第１の所定時間Ｔ１を超えなければ出力Ｚは０となって利得設定部８７が送話偏重モード設定用増幅器６の利得ＧＴ及び受話偏重モード設定用増幅器７の利得ＧＲを何れも０［ｄＢ］に設定することで中立モードに設定される。
【０１０７】
而して、回線側帰還利得βの推定値｜β’｜がしきい値β０以上であるときには他の条件にかかわらず偏重モード制御手段８が常に中立モードに設定し、回線側帰還利得βが相対的に大きい状況では受話偏重モードに設定しないため、送話ブロッキングの発生を抑えることができる。さらに、しきい値β０を設定可変なパラメータとして利用できるので、機器の設置環境に応じて機器の性能を様々に設定することが可能である。
【０１０８】
（実施形態１６）
本実施形態は、図１９に示すように基本的な構成が実施形態１１並びに実施形態１３と共通であるから共通の構成要素には同一の符号を付して説明を省略し、本実施形態の特徴となる構成についてのみ説明する。本実施形態では、音響側帰還利得推定手段１１と回線側帰還利得推定手段１２の両方を備え、送話信号音声区間検出部３７並びに受話信号音声区間検出部３８において、それぞれ回線側帰還利得βの推定値｜β’｜並びに音響側帰還利得αの推定値｜α’｜に応じて音声と非音声の判定に使用する送話側しきい値及び受話側しきい値Ｐs0，δが変更される点に特徴がある。なお、送話信号音声区間検出部３７及び受話信号音声区間検出部３８の構成は図２で示した実施形態１のものと共通であるから図示並びに説明は省略する。
【０１０９】
受話信号音声区間検出部３８には音響側帰還利得αの推定値｜α’｜が入力されており、この推定値｜α’｜が所定の基準値よりも小さければ、音声／非音声判定部にて瞬時パワー推定値Ｐｓと比較されるしきい値Ｐs0並びに瞬時パワー推定値Ｐｓと背景雑音パワー推定値Ｐｎとの比Ｐｓ／Ｐｎと比較されるしきい値δが予め決められた最適値に設定され、推定値｜α’｜が基準値以上であれば、これら２つのしきい値Ｐs0，δが上記最適値よりも小さい値に設定される。これにより、音響側帰還利得αが相対的に大きい状況下において受話信号音声区間検出部３８によって音声区間が検出されやすくなり、音声区間が検出されている場合に送話偏重モードに設定されないために受話ブロッキングが生じにくくなる。
【０１１０】
一方、送話信号音声区間検出部３７には回線側帰還利得βの推定値｜β’｜が入力されており、この推定値｜β’｜が所定の基準値よりも小さければ、音声／非音声判定部にて瞬時パワー推定値Ｐｓと比較されるしきい値Ｐs0並びに瞬時パワー推定値Ｐｓと背景雑音パワー推定値Ｐｎとの比Ｐｓ／Ｐｎと比較されるしきい値δが予め決められた最適値に設定され、推定値｜β’｜が基準値以上であれば、これら２つのしきい値Ｐs0，δが上記最適値よりも小さい値に設定される。これにより、回線側帰還利得βが相対的に大きい状況下において送話信号音声区間検出部３７によって音声区間が検出されやすくなり、音声区間が検出されている場合に受話偏重モードに設定されないために送話ブロッキングが生じにくくなる。
【０１１１】
なお、帰還利得α，βの推定値｜α’｜，｜β’｜に応じて２つのしきい値Ｐs0，δを変更する代わりに、各音声区間検出部３７，１０を構成する瞬時パワー推定部や背景雑音パワー推定部における演算で使用される時定数を変更するようにしても同様の効果を奏する。例えば、音響側帰還利得α，βの推定値｜α’｜，｜β’｜が所定の基準値よりも小さければ、時定数が予め決められた最適値に設定され、推定値｜α’｜，｜β’｜が基準値以上であれば、最適値に対して立ち上がり特性が遅く且つ立ち下がり特性が速くなるような値に時定数を設定すればよい。
【０１１２】
（実施形態１７）
図２０は本実施形態の音声切換装置ＶＳを備えた拡声通話端末の概略構成を示しており、実施形態７と共通の構成要素には同一の符号を付している。
【０１１３】
本実施形態における拡声通話端末では、実施形態６と同様に送話信号音声区間検出部３７に対する送話側しきい値及び送話側時定数と、受話信号音声区間検出部３８に対する受話側しきい値及び送話側時定数とが、音響側帰還利得αの推定値｜α’｜としきい値α０の大小関係、並びに回線側帰還利得βの推定値｜β’｜としきい値β０の大小関係に応じた複数種類（本実施形態では２種類）の値としてフラッシュメモリ１１０に格納している。
【０１１４】
本実施形態における拡声通話端末では、実施形態７と同様に、通話を開始する際にＣＰＵ２００からＤＳＰ１００に対して通話開始要求コマンドを通信用インタフェースを介して送信し、この通話開始要求コマンドを受け取ったコマンド処理手段１０３がその内容を解釈し、送話信号音声区間検出部３７並びに受話信号音声区間検出部３８において用いる上記パラメータ（送話側しきい値、送話側時定数、受話側しきい値、受話側時定数）の格納場所を示すアドレスデータをパラメータ設定手段１０４に与える。パラメータ設定手段１０４では受け取ったアドレスデータに基づいて送話側及び受話側のしきい値並びに時定数のデータをフラッシュメモリ１１０から読み込む。このとき、音響側帰還利得αの推定値｜α’｜としきい値α０の大小関係、並びに回線側帰還利得βの推定値｜β’｜としきい値β０の大小関係に応じた値のパラメータが選択され、パラメータ設定手段１０４により音声区間検出部３７，１０を実現するソフトウェアモジュールに対して初期設定される。これら初期設定が完了した後、パラメータ設定手段１０４はパラメータの初期化が完了したことを示すフラグをセットし、このフラグのセットを受けてコマンド処理手段１０３が応答コマンドをＣＰＵ２００に対して送信する。
【０１１５】
而して、本実施形態においては実施形態７と同様に近端側の伝達特性と遠端側の伝達特性の差を簡単に補正できるとともに送話ブロッキング並びに受話ブロッキングを生じにくくできる。
【０１１６】
（実施形態１８）
図２１は本実施形態の音声切換装置ＶＳを備えた拡声通話端末の概略構成を示しており、実施形態８と共通の構成要素には同一の符号を付している。
【０１１７】
本実施形態では、送話信号音声区間検出部３７に対する送話側しきい値及び送話側時定数と、受話信号音声区間検出部３８に対する受話側しきい値及び送話側時定数とが、音響側帰還利得αの推定値｜α’｜としきい値α０の大小関係、並びに回線側帰還利得βの推定値｜β’｜としきい値β０の大小関係に応じた複数種類（本実施形態では２種類）の値に設定されるだけでなく、相手側の通話端末の種類（例えば、ドアホンとロビーインターホン）に応じて各々異なる値に設定可能としている。具体的には、実施形態８と同様にフラッシュメモリ１１０の２つの領域Ｍ１，Ｍ２にそれぞれドアホン用のパラメータ（送話側及び受話側のしきい値並びに時定数）とロビーインターホン用のパラメータを格納している。そして、ＣＰＵ２００が呼出相手の通話端末の種類に応じたパラメータのアドレスをコマンドで指定し、パラメータ設定手段１０４が受け取ったアドレスデータに基づいて相手側の通話端末の種類に応じた送話側及び受話側のしきい値並びに時定数のデータをフラッシュメモリ１１０の領域Ｍ１，Ｍ２から読み込む。このとき、実施形態１７と同様に音響側帰還利得αの推定値｜α’｜としきい値α０の大小関係、並びに回線側帰還利得βの推定値｜β’｜としきい値β０の大小関係に応じた値のパラメータが選択され、パラメータ設定手段１０４により音声区間検出部３７，１０を実現するソフトウェアモジュールに対して初期設定される。
【０１１８】
而して、本実施形態においては送話信号音声区間検出部３７並びに受話信号音声区間検出部３８に対する送話側及び受話側のしきい値及び時定数が相手側の通話端末毎に外部（ＣＰＵ２００）から各々個別に設定可能であるから、相手側の通話端末毎に異なる２つの伝達特性の差を簡単に補正することができるとともに送話ブロッキング並びに受話ブロッキングを生じにくくできる。
【０１１９】
（実施形態１９）
本実施形態は、偏重モード制御手段８が、近端側雑音パワーの推定値ＰＮｎと遠端側雑音パワーの推定値ＰＦｎとの差と、音響側並びに回線側の帰還利得推定値｜α’｜，｜β’｜を参照して送話偏重モード設定用増幅器６並びに受話偏重モード設定用増幅器７の利得ＧＴ，ＧＲを適応的に更新する点に特徴がある。但し、全体の構成は実施形態１５と同一であるから図示並びに説明は省略する。
【０１２０】
偏重モード制御手段８は、送話偏重モードに設定する場合に近端側雑音パワーの推定値ＰＮｎと遠端側雑音パワーの推定値ＰＦｎとの差（＝ＰＦｎ−ＰＮｎ）が大きいほど送話偏重モード設定用増幅器６の利得ＧＴを大きくし、受話偏重モードに設定する場合に近端側雑音パワーの推定値ＰＮｎと遠端側雑音パワーの推定値ＰＦｎとの差（＝ＰＮｎ−ＰＦｎ）が大きいほど受話偏重モード設定用増幅器７の利得ＧＲを小さくするような制御を行う。ここで、送話偏重モード設定用増幅器６の利得ＧＴを増大するときは音響側帰還利得αの推定値｜α’｜がしきい値α０を下回っていることを条件とし、受話偏重モード設定用増幅器７の利得ＧＲを増大するときは回線側帰還利得βの推定値｜β’｜がしきい値β０を下回っていることを条件とする。さらに両利得ＧＴ，ＧＲの増減に応じて利得ＧＴとしきい値α０の積及び利得ＧＲとしきい値β０の積が何れもほぼ一定値となるように両しきい値α０，β０を増減させる。なお、両利得ＧＴ，ＧＲの値を更新する場合、送話信号及び受話信号の各音声区間検出部３７，３８が何れも音声区間を検出していないことを条件とする。
【０１２１】
上述のようにして偏重モード制御手段８が近端側雑音パワーの推定値ＰＮｎと遠端側雑音パワーの推定値ＰＦｎとの差と、音響側並びに回線側の帰還利得推定値｜α’｜，｜β’｜を参照して送話偏重モード設定用増幅器６並びに受話偏重モード設定用増幅器７の利得ＧＴ，ＧＲを適応的に更新するものであり、これにより、通話中の不自然な言葉の途切れを発生させることが無く、通話モードを送話モード又は受話モードにバランスよく切り換えることができ、さらに通話モードの切り換えに必要な遠端側及び近端側の発声レベルを下げることができる。
【０１２２】
ところで、上述のように偏重モード制御手段８により適応的に更新される送話偏重モード設定用増幅器６並びに受話偏重モード設定用増幅器７の利得ＧＴ，ＧＲには上限値を設定することが望ましい。すなわち、両利得ＧＴ，ＧＲが必要以上に大きな値に設定されると受話ブロッキングや送話ブロッキングが生じてしまうことになるから、利得ＧＴ，ＧＲに予め上限値を設定しておいて偏重モード制御手段８が両利得ＧＴ，ＧＲを適応的に更新する際に上限値を超えないようにすることで受話ブロッキングや送話ブロッキングの発生を抑えることができるようになる。
【０１２３】
【発明の効果】
請求項１の発明は、マイクロホン及びスピーカを有する拡声通話端末が他の通話端末又は拡声通話端末に有線で接続される拡声通話系の前記拡声通話端末に用いられ、前記マイクロホンで集音する送話信号を回線へ伝送するための送話側信号経路に損失を挿入する送話側損失挿入手段と、回線から受信した受話信号を前記スピーカへ伝送するための受話側信号経路に損失を挿入する受話側損失挿入手段と、前記送話側損失挿入手段に入力される送話信号を取り出して増幅する送話偏重モード設定用増幅手段と、前記受話側損失挿入手段に入力される受話信号を取り出して増幅する受話偏重モード設定用増幅手段と、前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段で増幅された送話信号及び受話信号に基づいて通話モードを推定するとともに該推定結果に応じて前記送話側損失挿入手段並びに受話側損失挿入手段が前記経路に挿入する損失量を制御して前記通話モードを送話モードと受話モードに切り換える挿入損失量制御手段と、送話信号に含まれる近端側の雑音パワーを推定する近端側雑音パワー推定手段と、受話信号に含まれる遠端側の雑音パワーを推定する遠端側雑音パワー推定手段と、遠端側雑音パワー並びに近端側雑音パワーの各推定値に応じて前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の各利得を調整する偏重モード制御手段とを備え、前記挿入損失量制御手段が、前記送話側損失挿入手段への入力信号の瞬時パワーを推定する第１の瞬時パワー推定部と、前記受話側損失挿入手段への入力信号の瞬時パワーを推定する第２の瞬時パワー推定部と、前記送話側損失挿入手段への入力点から前記送話側損失挿入手段並びに回線側での回り込みを経て前記受話側損失挿入手段への入力点へ帰還する系の利得に応じて決定される値を係数にもつ回線帰還利得乗算部と、前記受話側損失挿入手段への入力点から前記受話側損失挿入手段並びに音響側での回り込みを経て前記送話側損失挿入手段への入力点へ到る経路の利得に応じて決定される値を係数にもつ音響帰還利得乗算部と、第２の瞬時パワー推定部の出力信号を音響帰還利得乗算部へ入力して得られる出力信号と第１の瞬時パワー推定部の出力信号との大小関係を比較する第１の比較器と、第１の瞬時パワー推定部の出力信号を回線帰還利得乗算部へ入力して得られる出力信号と第２の瞬時パワー推定部の出力信号との大小関係を比較する第２の比較器と、送話信号の音声区間を検出する送話信号音声区間検出部と、受話信号の音声区間を検出する受話信号音声区間検出部と、第１の比較器及び第２の比較器の比較結果と送話信号音声区間検出部及び受話信号音声区間検出部の検出結果とに基づいて通話状態を判定し送話側損失挿入手段及び受話側損失挿入手段の挿入損失量を制御する挿入損失量分配処理部とを具備し、前記偏重モード制御手段は、遠端側並びに近端側の各雑音パワーの推定値の大小関係と前記送話信号音声区間検出部の検出結果に基づいて、前記送話偏重モード設定用増幅手段及び前記受話偏重モード設定用増幅手段の各利得をほぼ等しくする中立モードと前記受話偏重モード設定用増幅手段の利得を前記送話偏重モード設定用増幅手段の利得よりも増大させる受話偏重モードの何れかに設定することを特徴とし、従来のように近端側雑音パワー並びに遠端側雑音パワーの各推定値のみに基づいて偏重モードを設定するのではなく、遠端側並びに近端側の各雑音パワーの推定値の大小関係と送話信号音声区間検出部の検出結果に基づいて偏重モード制御手段が偏重モードを設定するのであり、例えば近端側から一方的に発声している場合に偏重モード制御手段が受話偏重モードに設定しないから、近端側の話者が発声しているにもかかわらず挿入損失量制御手段が送話モードから受話モードに誤って切り換えてしまうのを防止することができ、音声の途切れや不自然な抑揚を抑えることができる。
【０１２４】
請求項２の発明は、請求項１の発明において、前記偏重モード制御手段は、前記近端側雑音パワーの推定値が前記遠端側雑音パワーの推定値と所定の第１の雑音パワー比係数との積以上の値となり且つ前記送話信号音声区間検出部によって音声区間が検出されない状態の継続時間が第１の所定時間以上となったときに受話偏重モード設定用増幅手段の利得を送話偏重モード設定用増幅手段の利得よりも増大させて受話偏重モードに設定することを特徴とし、第１の雑音パワー比係数と第１の所定時間によって偏重モード制御手段による受話偏重モードへの切換特性を任意に設定することができる。
【０１２５】
請求項３の発明は、上記目的を達成するために、マイクロホン及びスピーカを有する拡声通話端末が他の通話端末又は拡声通話端末に有線で接続される拡声通話系の前記拡声通話端末に用いられ、前記マイクロホンで集音する送話信号を回線へ伝送するための送話側信号経路に損失を挿入する送話側損失挿入手段と、回線から受信した受話信号を前記スピーカへ伝送するための受話側信号経路に損失を挿入する受話側損失挿入手段と、前記送話側損失挿入手段に入力される送話信号を取り出して増幅する送話偏重モード設定用増幅手段と、前記受話側損失挿入手段に入力される受話信号を取り出して増幅する受話偏重モード設定用増幅手段と、前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段で増幅された送話信号及び受話信号に基づいて通話モードを推定するとともに該推定結果に応じて前記送話側損失挿入手段並びに受話側損失挿入手段が前記経路に挿入する損失量を制御して前記通話モードを送話モードと受話モードに切り換える挿入損失量制御手段と、送話信号に含まれる近端側の雑音パワーを推定する近端側雑音パワー推定手段と、受話信号に含まれる遠端側の雑音パワーを推定する遠端側雑音パワー推定手段と、遠端側雑音パワー並びに近端側雑音パワーの各推定値に応じて前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の各利得を調整する偏重モード制御手段とを備え、前記挿入損失量制御手段が、前記送話側損失挿入手段への入力信号の瞬時パワーを推定する第１の瞬時パワー推定部と、前記受話側損失挿入手段への入力信号の瞬時パワーを推定する第２の瞬時パワー推定部と、前記送話側損失挿入手段への入力点から前記送話側損失挿入手段並びに回線側での回り込みを経て前記受話側損失挿入手段への入力点へ帰還する系の利得に応じて決定される値を係数にもつ回線帰還利得乗算部と、前記受話側損失挿入手段への入力点から前記受話側損失挿入手段並びに音響側での回り込みを経て前記送話側損失挿入手段への入力点へ到る経路の利得に応じて決定される値を係数にもつ音響帰還利得乗算部と、第２の瞬時パワー推定部の出力信号を音響帰還利得乗算部へ入力して得られる出力信号と第１の瞬時パワー推定部の出力信号との大小関係を比較する第１の比較器と、第１の瞬時パワー推定部の出力信号を回線帰還利得乗算部へ入力して得られる出力信号と第２の瞬時パワー推定部の出力信号との大小関係を比較する第２の比較器と、送話信号の音声区間を検出する送話信号音声区間検出部と、受話信号の音声区間を検出する受話信号音声区間検出部と、第１の比較器及び第２の比較器の比較結果と送話信号音声区間検出部及び受話信号音声区間検出部の検出結果とに基づいて通話状態を判定し送話側損失挿入手段及び受話側損失挿入手段の挿入損失量を制御する挿入損失量分配処理部とを具備し、前記偏重モード制御手段は、遠端側並びに近端側の各雑音パワーの推定値の大小関係と前記受話信号音声区間検出部の検出結果に基づいて、前記送話偏重モード設定用増幅手段及び前記受話偏重モード設定用増幅手段の各利得をほぼ等しくする中立モードと前記送話偏重モード設定用増幅手段の利得を前記受話偏重モード設定用増幅手段の利得よりも増大させる送話偏重モードの何れかに設定することを特徴とし、従来のように近端側雑音パワー並びに遠端側雑音パワーの各推定値のみに基づいて偏重モードを設定するのではなく、遠端側並びに近端側の各雑音パワーの推定値の大小関係と受話信号音声区間検出部の検出結果に基づいて偏重モード制御手段が偏重モードを設定するのであり、例えば遠端側から一方的に発声している場合に偏重モード制御手段が送話偏重モードに設定しないから、遠端側の話者が発声しているにもかかわらず挿入損失量制御手段が受話モードから送話モードに誤って切り換えてしまうのを防止することができ、音声の途切れや不自然な抑揚を抑えることができる。
【０１２６】
請求項４の発明は、請求項３の発明において、前記偏重モード制御手段は、前記遠端側雑音パワーの推定値が前記近端側雑音パワーの推定値と所定の第２の雑音パワー比係数との積以上の値となり且つ前記受話信号音声区間検出部によって音声区間が検出されない状態の継続時間が第２の所定時間以上となったときに送話偏重モード設定用増幅手段の利得を受話偏重モード設定用増幅手段の利得よりも増大させて送話偏重モードに設定することを特徴とし、第２の雑音パワー比係数と第２の所定時間によって偏重モード制御手段による送話偏重モードへの切換特性を任意に設定することができる。
【０１２７】
請求項５の発明は、請求項１〜４の何れかの発明において、前記送話側又は受話側の音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記しきい値並びに前記所定値を外部から設定可能としたことを特徴とし、種々の拡声通話端末に容易に適応させることができる。
【０１２８】
請求項６の発明は、請求項１〜４の何れかの発明において、前記送話側又は受話側の音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記瞬時パワー推定部は立ち上がりが急峻で立ち下がりが緩やかな特性を有するフィルタからなり、前記雑音パワー推定部は立ち上がりが緩やかで立ち下がりが急峻な特性を有するフィルタからなり、該フィルタの前記特性を決定するパラメータを外部から設定可能としたことを特徴とし、種々の拡声通話端末に容易に適応させることができる。
【０１２９】
請求項７の発明は、請求項５又は６の発明において、前記しきい値及び所定値又は前記パラメータを前記送話側又は受話側の音声区間検出部に対して外部から個別に設定可能としたことを特徴とし、近端側の伝達特性と遠端側の伝達特性との差が簡単に補正できる。
【０１３０】
請求項８の発明は、請求項５又は６の発明において、前記しきい値及び所定値又は前記パラメータを前記他の通話端末の種類に応じて各々異なる値に設定可能としたことを特徴とし、相手側の通話端末毎に異なる近端側の伝達特性と遠端側の伝達特性との差が簡単に補正できる。
【０１３１】
請求項９の発明は、請求項１〜８の何れかの発明において、前記送話側又は受話側の音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記近端側雑音パワー推定手段を前記送話側音声区間検出部を構成する前記雑音パワー推定部で兼用するとともに、前記遠端側雑音パワー推定手段を前記受話側音声区間検出部を構成する前記雑音パワー推定部で兼用したことを特徴とし、構成が簡略化できる。
【０１３２】
請求項１０の発明は、請求項２の発明において、前記偏重モード制御手段は、前記継続時間が前記第１の所定時間未満であっても、前記近端側雑音パワーの推定値が前記遠端側雑音パワーの推定値と前記第１の雑音パワー比係数よりも大きい所定の第３の雑音パワー比係数との積以上となる状態が前記第１の所定時間よりも長い所定の第３の所定時間以上継続したときには受話偏重モードに設定することを特徴とし、近端側の通話端末の周囲に高レベルの非定常騒音が存在する場合でも受話偏重モードに設定され難くなるのを防いで通話モードが送話モードに固定されてしまう、所謂片倒れの発生が防止できる。
【０１３３】
請求項１１の発明は、請求項４の発明において、前記偏重モード制御手段は、前記継続時間が前記第２の所定時間未満であっても、前記遠端側雑音パワーの推定値が前記近端側雑音パワーの推定値と前記第２の雑音パワー比係数よりも大きい所定の第４の雑音パワー比係数との積以上となる状態が前記第２の所定時間よりも長い所定の第４の所定時間以上継続したときには送話偏重モードに設定することを特徴とし、遠端側の通話端末の周囲に高レベルの非定常騒音が存在する場合でも送話偏重モードに設定され難くなるのを防いで通話モードが受話モードに固定されてしまう、所謂片倒れの発生が防止できる。
【０１３４】
請求項１２の発明は、請求項３の発明において、前記受話側損失挿入手段の出力点から近端側の音響エコー経路を介して前記送話側損失挿入手段の入力点へ帰還する経路の音響側帰還利得を推定する音響側帰還利得推定手段を備え、前記偏重モード制御手段は、音響側帰還利得の推定値が所定の条件を満たさなければ送話偏重モードに移行しないことを特徴とし、受話ブロッキングを防止することができる。
【０１３５】
請求項１３の発明は、請求項１２の発明において、前記偏重モード制御手段は、前記遠端側雑音パワーの推定値が前記近端側雑音パワーの推定値と前記第２の雑音パワー比係数との積以上であり、前記受話信号音声区間検出部の検出結果が非音声区間であり、且つ前記音響側帰還利得の推定値が所定のしきい値未満である状態が所定時間以上継続したときに送話偏重モードに設定することを特徴とし、音響側帰還利得の推定値が所定のしきい値以上であるときには他の条件にかかわらず偏重モード制御手段が常に中立モードに設定し、音響側帰還利得が相対的に大きい状況では送話偏重モードに設定しないため、受話ブロッキングが防止できる。また、しきい値が設定可変なパラメータとして利用できるので、音声切換装置を利用する機器の設置環境に応じて当該機器の性能を様々に設定することが可能である。
【０１３６】
請求項１４の発明は、請求項１の発明において、前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たさなければ受話偏重モードに移行しないことを特徴とし、送話ブロッキングを防止することができる。
【０１３７】
請求項１５の発明は、請求項１４の発明において、前記偏重モード制御手段は、前記近端側雑音パワーの推定値が前記遠端側雑音パワーの推定値と前記第１の雑音パワー比係数との積以上であり、前記送話信号音声区間検出部の検出結果が非音声区間であり、且つ前記回線側帰還利得の推定値が所定のしきい値未満である状態が所定時間以上継続したときに受話偏重モードに設定することを特徴とし、回線側帰還利得の推定値が所定のしきい値以上であるときには他の条件にかかわらず偏重モード制御手段が常に中立モードに設定し、回線側帰還利得が相対的に大きい状況では受話偏重モードに設定しないため、送話ブロッキングが防止できる。また、しきい値が設定可変なパラメータとして利用できるので、音声切換装置を利用する機器の設置環境に応じて当該機器の性能を様々に設定することが可能である。
【０１３８】
請求項１６の発明は、請求項１２の発明において、前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たさなければ受話偏重モードに移行せず、前記送話側及び受話側の各音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記送話側及び受話側の各音声区間検出部は、前記音響側並びに回線側の各帰還利得推定手段の推定値に応じて前記しきい値及び所定値が変更されることを特徴とし、帰還利得が相対的に大きい状況下において音声区間検出部によって音声区間が検出されやすくなり、音声区間が検出されている場合に偏重モードに設定されないために送話あるいは受話ブロッキングが生じにくくなる。
【０１３９】
請求項１７の発明は、請求項１２の発明において、前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たす場合には受話偏重モードに設定し、前記送話側及び受話側の各音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記瞬時パワー推定部は立ち上がりが急峻で立ち下がりが緩やかな特性を有するフィルタからなり、前記雑音パワー推定部は立ち上がりが緩やかで立ち下がりが急峻な特性を有するフィルタからなり、前記送話側及び受話側の各音声区間検出部は、前記音響側並びに回線側の各帰還利得推定手段の推定値に応じて前記特性を決定するパラメータが変更されることを特徴とし、帰還利得が相対的に大きい状況下において音声区間検出部によって音声区間が検出されやすくなり、音声区間が検出されている場合に偏重モードに設定されないために送話あるいは受話ブロッキングが生じにくくなる。
【０１４０】
請求項１８の発明は、請求項１７の発明において、前記しきい値及び所定値と前記パラメータを前記送話側並びに受話側の各音声区間検出部に対して外部から個別に設定可能としたことを特徴とし、近端側の伝達特性と遠端側の伝達特性との差が簡単に補正できるとともに送話あるいは受話ブロッキングが生じにくくなる。
【０１４１】
請求項１９の発明は、請求項１７の発明において、前記しきい値及び所定値と前記パラメータを前記他の通話端末の種類に応じて各々異なる値に設定可能としたことを特徴とし、相手側の通話端末毎に異なる近端側と遠端側の２つの伝達特性の差を簡単に補正することができるとともに送話ブロッキング並びに受話ブロッキングを生じにくくできる。
【０１４２】
請求項２０の発明は、請求項１２の発明において、前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たさなければ受話偏重モードに移行せず、前記近端側雑音パワーの推定値と前記遠端側雑音パワーの推定値との差と、前記音響側並びに回線側の帰還利得推定値を参照して前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の利得を適応的に更新することを特徴とし、通話中の不自然な言葉の途切れを発生させることが無く、通話モードを送話モード又は受話モードにバランスよく切り換えることができ、さらに通話モードの切り換えに必要な遠端側及び近端側の発声レベルを下げることができる。
【０１４３】
請求項２１の発明は、請求項２０の発明において、前記偏重モード制御手段により適応的に更新される前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の利得に上限値を設定したことを特徴とし、偏重モード制御手段が両利得を適応的に更新する際に上限値を超えないようにすることで受話ブロッキングや送話ブロッキングの発生を抑えることができる。
【図面の簡単な説明】
【図１】実施形態１を示すブロック図である。
【図２】同上における送話信号音声区間検出手段のブロック図である。
【図３】同上における偏重モード制御手段の動作を説明するためのフローチャートである。
【図４】実施形態２における偏重モード制御手段のブロック図である。
【図５】実施形態３を示すブロック図である。
【図６】同上における偏重モード制御手段の動作を説明するためのフローチャートである。
【図７】実施形態４における偏重モード制御手段のブロック図である。
【図８】実施形態５の音声切換装置を備えた拡声通話端末を示す一部省略したブロック図である。
【図９】実施形態６の音声切換装置を備えた拡声通話端末を示す一部省略したブロック図である。
【図１０】実施形態７の音声切換装置を備えた拡声通話端末を示す一部省略したブロック図である。
【図１１】実施形態８の音声切換装置を備えた拡声通話端末を示す一部省略したブロック図である。
【図１２】実施形態９を示すブロック図である。
【図１３】実施形態１０における偏重モード制御手段のブロック図である。
【図１４】実施形態１１における偏重モード制御手段のブロック図である。
【図１５】実施形態１２を示すブロック図である。
【図１６】実施形態１３における偏重モード制御手段のブロック図である。
【図１７】実施形態１４を示すブロック図である。
【図１８】実施形態１５における偏重モード制御手段のブロック図である。
【図１９】実施形態１６を示すブロック図である。
【図２０】実施形態１７の音声切換装置を備えた拡声通話端末を示す一部省略したブロック図である。
【図２１】実施形態１８の音声切換装置を備えた拡声通話端末を示す一部省略したブロック図である。
【図２２】従来例を示すブロック図である。
【符号の説明】
１ 送話側損失挿入手段
２ 受話側損失挿入手段
３ 挿入損失量制御手段
４ 近端側雑音パワー推定手段
５ 遠端側雑音パワー推定手段
６ 送話偏重モード設定用増幅器
７ 受話偏重モード設定用増幅器
８ 偏重モード制御手段
３７ 送話信号音声区間検出部[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a voice switching device used in a loudspeaker terminal used in an intercom system of a housing complex.
[0002]
[Prior art]
Conventionally, it is not necessary to have a handset during a call, and a voice signal transmitted from the other party's call terminal is sent to the caller away from the call terminal by a speaker, and the voice emitted by the caller is microphone A loudspeaker call system is provided that enables half-duplex calling by collecting sound and transmitting it to the other party's call terminal. In such a loudspeaker communication system, a two-wire / four-wire conversion hybrid circuit required when acoustic coupling between a speaker and a microphone, which are constituent elements thereof, and a transmission path of an audio signal are configured in a two-wire form. A closed loop is formed on the speech path by wraparound from the transmission signal path to the reception signal path caused by impedance mismatch in the voice, and acoustic coupling between the speaker and the microphone in the other party's speech terminal. If it becomes more than twice, howling occurs, and if howling occurs, the call cannot be continued, and means for suppressing this is required.
[0003]
Therefore, conventionally, it is determined whether the call state is the reception state or the transmission state by monitoring the transmission signal and the reception signal, and the transmission signal path or the reception signal path is determined according to the determined communication state. A voice switching device (so-called voice switch) that prevents a howling by reducing a loop gain of a closed loop by inserting a loss into at least one of them has been widely used in a voice call terminal. The basic operation of the voice switching device is to estimate the powers of the transmission signal and the reception signal, and to compare the magnitude relationship between them, and to insert a predetermined amount of loss on the side having a smaller instantaneous power.
[0004]
FIG. 22 is a block diagram showing a conventional voice switching device disclosed in Patent Document 1. In FIG. This conventional example includes transmission side loss insertion means 1 for inserting a loss into a transmission side signal path L1 for transmitting a transmission signal collected by a microphone (not shown) of a voice communication terminal to the line, and a line. Is input to the reception side loss insertion means 2 for inserting a loss into the reception side signal path L2 for transmitting the reception signal received from the speaker to the loudspeaker terminal speaker (not shown) and the transmission side loss insertion means 1. A transmission bias mode setting amplifier 6 for extracting and amplifying a transmission signal, a reception bias mode setting amplifier 7 for extracting and amplifying a reception signal input to the reception side loss insertion means 2, and a transmission bias mode setting The speech mode is estimated based on the transmission signal and the reception signal amplified by the amplifier 6 and the reception bias mode setting amplifier 7, and the transmission side loss insertion means 1 and the reception side loss insertion means 2 correspond to the estimation result. Insertion loss amount control means 3 for switching the speech mode to the transmission mode, the reception mode and the neutral mode by controlling the loss amount inserted in the paths L1 and L2, and the near-end side noise power included in the transmission signal are estimated. Near-end side noise power estimation means 4, far-end side noise power estimation means 5 for estimating the far-end side noise power included in the received signal, and far-end side noise power and estimated values PFn of the near-end side noise power , And a deviation mode control means 8 ′ for adjusting the gains of the transmission deviation mode setting amplifier 6 and the reception deviation mode setting amplifier 7 according to PNn.
[0005]
The far-end side noise power estimation means 5 and the near-end side noise power estimation means 4 are both realized by an integration circuit or a digital filter having a characteristic that the rise is gradual and the fall is steep. The estimation means 5 estimates the background noise power that is constantly present in the received signal, and the near-end noise power estimation means 4 estimates the noise power that is constantly present in the transmission signal.
[0006]
If the estimated value PFn of the far-end side noise power is sufficiently larger than the estimated value PNn of the near-end side noise power (PFn >> PNn), the eccentric mode control unit 8 ′ transmits the transmission eccentricity mode setting amplifier 6. Is set to G (> 0) [dB], and the gain GR of the reception bias mode setting amplifier 7 is set to 0 [dB], so that the call mode is set to the transmission bias mode and the near-end side noise power is estimated. If the value PNn is sufficiently larger than the estimated value PFn of the far-end side noise power (PNn >> PFn), the gain GR of the reception weighting mode setting amplifier 7 is set to G [dB], and the transmission weighting mode setting amplifier By setting the gain GT of 6 to 0 [dB], the call mode is set to the reception bias mode, and the difference between the far-end side noise power estimation value PFn and the near-end side noise power estimation value PNn is sufficiently large. If there is not, for setting the listening bias mode The gains GR and GT of the amplifier 7 and the transmission bias mode setting amplifier 6 are set to 0 [dB] and set to the neutral mode.
[0007]
That is, when the difference between the ambient noise level on the far end side and the ambient noise level on the near end side is large, the insertion loss amount control means 3 that monitors the transmission signal and the received signal to estimate the call state, for example, In the situation where the ambient noise level on the side is large, it is always determined as the reception state, and in the situation where the near-end ambient noise level is large, it is always determined as the transmission state, regardless of the actual call state. There may be a phenomenon (so-called one-sided fall) that fixes the call state in one of the transmission states.
[0008]
On the other hand, in the conventional example, as described above, the bias mode control means 8 ′ compares the estimated value PFn of the far-end side noise power with the estimated value PNn of the near-end side noise power, and estimates the far-end side noise power. If the value PFn is sufficiently larger, the insertion loss amount control means 3 amplifies the transmission signal monitored by the insertion loss amount control means 3 by the transmission bias mode setting amplifier 6 by a gain G [dB]. When the transmission state is easily determined (transmission eccentric mode), on the contrary, when the near-end noise power estimation value PNn is sufficiently larger, the reception signal monitored by the insertion loss amount control means 3 is received. By amplifying only the gain G [dB] by the bias mode setting amplifier 7, the insertion loss amount control means 3 is set to a state in which it is easy to determine the reception state (reception bias mode), thereby suppressing the above-mentioned fall-down. Good switching characteristics can be obtained. And can be.
[0009]
[Patent Document 1]
JP 2002-359580 A (paragraph 0081-paragraph 0085, FIG. 21)
[0010]
[Problems to be solved by the invention]
However, in the above conventional example, for example, when unilaterally speaking from the near end side, the estimated value PNn of the near end side noise power gradually increases, and the difference between the ambient noise levels on the near end side and the far end side is increased. Is set to the reception bias mode by the bias mode control means 8 'even if the signal is small, the insertion loss amount control means 3 erroneously switches from the transmission mode to the reception mode even though the near-end speaker is speaking. The sound may be interrupted or unnatural inflection may occur. Similarly, when the voice is unilaterally spoken from the far end side, the far end side noise power estimation value PFn gradually increases and is set to the transmission bias mode by the bias mode control means 8 '. However, the insertion loss amount control means 3 erroneously switches from the reception mode to the transmission mode, and the voice is interrupted or unnatural inflection may occur.
[0011]
The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a voice switching device that can prevent voice interruption and inflection by preventing an erroneous call mode from being set. .
[0012]
[Means for Solving the Problems]
In order to achieve the above object, the invention according to claim 1 is used for the above-mentioned loudspeaker terminal of a loudspeaker system in which a loudspeaker terminal having a microphone and a speaker is connected to another telephone terminal or a loudspeaker terminal by wire, Transmission side loss insertion means for inserting a loss into a transmission side signal path for transmitting a transmission signal collected by the microphone to the line, and a reception side for transmitting a reception signal received from the line to the speaker A receiver-side loss insertion unit that inserts a loss into a signal path; a transmission bias mode setting amplification unit that extracts and amplifies a transmission signal input to the transmission-side loss insertion unit; and the reception-side loss insertion unit. A receiving bias mode setting amplifying means for picking up and amplifying the received receiving signal; a transmission bias mode setting amplifying means; and a transmission signal amplified by the receiving bias mode setting amplifying means. Estimating the call mode based on the received signal and controlling the amount of loss inserted into the path by the transmission side loss insertion means and the reception side loss insertion means in accordance with the estimation result, thereby changing the call mode to the transmission mode. Insertion loss control means for switching to the reception mode, near-end noise power estimation means for estimating the near-end noise power included in the transmitted signal, and far-end noise power for estimating the far-end noise power included in the received signal End-side noise power estimation means, and a bias that adjusts the gains of the transmission bias mode setting amplification means and the reception bias mode setting amplification means according to the estimated values of the far-end side noise power and the near-end side noise power Mode control means, wherein the insertion loss amount control means estimates the instantaneous power of the input signal to the transmission side loss insertion means, and the reception side loss insertion means A second instantaneous power estimator for estimating the instantaneous power of the input signal of the receiver, and the insertion of the receiving side loss from the input point to the transmitting side loss insertion means through the transmission side loss insertion means and the wraparound on the line side A line feedback gain multiplier having as a coefficient a value determined according to the gain of the system that feeds back to the input point to the input unit, and from the input point to the reception side loss insertion unit to the reception side loss insertion unit and the acoustic side And an acoustic feedback gain multiplier having a value determined according to the gain of the path reaching the input point to the transmission side loss insertion means via the wraparound and the output signal of the second instantaneous power estimator The first comparator for comparing the magnitude relationship between the output signal obtained by input to the acoustic feedback gain multiplication unit and the output signal of the first instantaneous power estimation unit, and the output signal of the first instantaneous power estimation unit Output signal obtained by input to feedback gain multiplier A second comparator for comparing the magnitude relationship between the signal and the output signal of the second instantaneous power estimation unit, a transmission signal speech segment detection unit for detecting a speech segment of the transmitted signal, and a speech segment of the received signal. The call state is determined based on the detected reception signal voice interval detection unit, the comparison results of the first comparator and the second comparator, and the detection results of the transmission signal voice interval detection unit and the reception signal voice interval detection unit. An insertion loss amount distribution processing unit for controlling the insertion loss amount of the transmission side loss insertion means and the reception side loss insertion means, and the bias mode control means is configured to control each noise power of the far end side and the near end side. Based on the magnitude relationship of the estimated values and the detection result of the transmission signal voice section detection unit, the neutral mode for substantially equalizing the gains of the transmission bias mode setting amplification unit and the reception bias mode setting amplification unit, and Of the amplification means for setting the reception bias mode The resulting than the gain of the transmission unbalance mode setting means for amplifying and setting either of the receiving overemphasis modes increase.
[0013]
According to a second aspect of the present invention, in the first aspect of the invention, the bias mode control means is configured such that the estimated value of the near-end side noise power is equal to the estimated value of the far-end side noise power and a predetermined first noise power ratio coefficient. When the duration of the state in which no speech section is detected by the transmission signal speech section detecting unit is equal to or longer than the first predetermined time, the gain of the reception bias mode setting amplifier is transmitted. The reception mode is set to be greater than the gain of the bias mode setting amplification means.
[0014]
In order to achieve the above object, the invention according to claim 3 is used for the above-mentioned loudspeaker call terminal of a loudspeaker call system in which a loudspeaker call terminal having a microphone and a speaker is connected to another call terminal or a loudspeaker call terminal by wire, Transmission side loss insertion means for inserting a loss into a transmission side signal path for transmitting a transmission signal collected by the microphone to the line, and a reception side for transmitting a reception signal received from the line to the speaker A receiver-side loss insertion unit that inserts a loss into a signal path; a transmission bias mode setting amplification unit that extracts and amplifies a transmission signal input to the transmission-side loss insertion unit; and the reception-side loss insertion unit. A receiving bias mode setting amplifying means for picking up and amplifying the received receiving signal; a transmission bias mode setting amplifying means; and a transmission signal amplified by the receiving bias mode setting amplifying means. Estimating the call mode based on the received signal and controlling the amount of loss inserted into the path by the transmission side loss insertion means and the reception side loss insertion means in accordance with the estimation result, thereby changing the call mode to the transmission mode. Insertion loss control means for switching to the reception mode, near-end noise power estimation means for estimating the near-end noise power included in the transmitted signal, and far-end noise power for estimating the far-end noise power included in the received signal End-side noise power estimation means, and a bias that adjusts the gains of the transmission bias mode setting amplification means and the reception bias mode setting amplification means according to the estimated values of the far-end side noise power and the near-end side noise power Mode control means, wherein the insertion loss amount control means estimates the instantaneous power of the input signal to the transmission side loss insertion means, and the reception side loss insertion means A second instantaneous power estimator for estimating the instantaneous power of the input signal of the receiver, and the insertion of the receiving side loss from the input point to the transmitting side loss insertion means through the transmission side loss insertion means and the wraparound on the line side A line feedback gain multiplier having as a coefficient a value determined according to the gain of the system that feeds back to the input point to the input unit, and from the input point to the reception side loss insertion unit to the reception side loss insertion unit and the acoustic side And an acoustic feedback gain multiplier having a value determined according to the gain of the path reaching the input point to the transmission side loss insertion means via the wraparound and the output signal of the second instantaneous power estimator The first comparator for comparing the magnitude relationship between the output signal obtained by input to the acoustic feedback gain multiplication unit and the output signal of the first instantaneous power estimation unit, and the output signal of the first instantaneous power estimation unit Output signal obtained by input to feedback gain multiplier A second comparator for comparing the magnitude relationship between the signal and the output signal of the second instantaneous power estimation unit, a transmission signal speech segment detection unit for detecting a speech segment of the transmitted signal, and a speech segment of the received signal. The call state is determined based on the detected reception signal voice interval detection unit, the comparison results of the first comparator and the second comparator, and the detection results of the transmission signal voice interval detection unit and the reception signal voice interval detection unit. An insertion loss amount distribution processing unit for controlling the insertion loss amount of the transmission side loss insertion means and the reception side loss insertion means, and the bias mode control means is configured to control each noise power of the far end side and the near end side. Based on the magnitude relationship of the estimated values and the detection result of the reception signal voice section detection unit, the neutral mode for substantially equalizing the gains of the transmission bias mode setting amplification unit and the reception bias mode setting amplification unit and the transmission mode Of the amplification means for setting the talk bias mode The resulting than the gain of the receiver unbalance mode setting means for amplifying and setting to one of the transmission overemphasis modes increase.
[0015]
According to a fourth aspect of the present invention, in the third aspect of the present invention, the bias mode control means is configured such that the estimated value of the far-end side noise power is equal to the estimated value of the near-end side noise power and a predetermined second noise power ratio coefficient. And the gain of the transmission bias mode setting amplification means when the duration of the state in which no voice section is detected by the reception signal voice section detector becomes a second predetermined time or more. It is characterized in that the transmission bias mode is set to be larger than the gain of the mode setting amplification means.
[0016]
According to a fifth aspect of the present invention, in the invention according to any one of the first to fourth aspects, the transmission side or reception side voice section detection unit is included in an input signal to the transmission side or reception side loss insertion means. A noise power estimator for estimating noise power; and comparing the first or second instantaneous power estimated value with a predetermined threshold value and a product of the noise power estimated value and the predetermined value. A speech / non-speech determination unit that determines that the input signal to the loss insertion means is a speech signal when it is greater than the threshold and the instantaneous power estimation value is greater than the product, and determines that it is a non-speech signal when it is not greater The threshold value and the predetermined value can be set from the outside.
[0017]
The invention of claim 6 is the invention according to any one of claims 1 to 4, wherein the voice section detecting unit on the transmitting side or the receiving side is included in an input signal to the loss insertion means on the transmitting side or the receiving side. A noise power estimator for estimating noise power; and comparing the first or second instantaneous power estimated value with a predetermined threshold value and a product of the noise power estimated value and the predetermined value. A speech / non-speech determination unit that determines that the input signal to the loss insertion means is a speech signal when it is greater than the threshold and the instantaneous power estimation value is greater than the product, and determines that it is a non-speech signal when it is not greater And the instantaneous power estimator comprises a filter having a steep rise and a gradual fall, and the noise power estimator has a slow rise and a steep fall. Consist filter having, characterized in that the settable parameters that determine the characteristics of the filter from the outside.
[0018]
The invention of claim 7 is the invention of claim 5 or 6, wherein the threshold value and the predetermined value or the parameter can be individually set from the outside to the voice section detecting unit on the transmitting side or the receiving side. It is characterized by that.
[0019]
The invention of claim 8 is characterized in that, in the invention of claim 5 or 6, the threshold value and the predetermined value or the parameter can be set to different values depending on the type of the other call terminal. .
[0020]
The invention according to claim 9 is the invention according to any one of claims 1 to 8, wherein the voice section detecting unit on the transmitting side or the receiving side is included in an input signal to the loss insertion means on the transmitting side or the receiving side. A noise power estimator for estimating noise power; and comparing the first or second instantaneous power estimated value with a predetermined threshold value and a product of the noise power estimated value and the predetermined value. A speech / non-speech determination unit that determines that the input signal to the loss insertion means is a speech signal when it is greater than the threshold and the instantaneous power estimation value is greater than the product, and determines that it is a non-speech signal when it is not greater The near-end side noise power estimation means is also used as the noise power estimation section constituting the transmission-side voice section detection section, and the far-end side noise power estimation means is used as the reception-side voice section detection. Characterized by being combined with said noise power estimating section constituting a part.
[0021]
According to a tenth aspect of the present invention, in the second aspect of the present invention, the deviation mode control means is configured such that the estimated value of the near-end side noise power is the far end even if the duration is less than the first predetermined time. A predetermined third predetermined condition in which a state that is equal to or greater than a product of the estimated value of the side noise power and a predetermined third noise power ratio coefficient larger than the first noise power ratio coefficient is longer than the first predetermined time. It is characterized in that it is set to the reception bias mode when it continues for more than a time.
[0022]
According to an eleventh aspect of the present invention, in the invention according to the fourth aspect, the deviation mode control means has the estimated value of the far end side noise power as the near end even if the duration is less than the second predetermined time. A predetermined fourth predetermined condition in which a state in which the estimated value of the side noise power and a predetermined fourth noise power ratio coefficient larger than the second noise power ratio coefficient is equal to or greater than the second predetermined time is longer It is characterized in that it is set to the transmission bias mode when it continues for more than a time.
[0023]
The invention of claim 12 is claimed in claim 3 The acoustic side feedback gain for estimating the acoustic side feedback gain of the path returning from the output point of the reception side loss insertion means to the input point of the transmission side loss insertion means via the acoustic echo path on the near end side The deviation mode control means includes an estimation unit, and the estimated value of the acoustic feedback gain does not shift to the transmission deviation mode unless the estimated value of the acoustic side feedback gain satisfies a predetermined condition.
[0024]
According to a thirteenth aspect of the present invention, in the twelfth aspect of the invention, the bias mode control means is configured such that the estimated value of the far-end side noise power is an estimated value of the near-end side noise power and the second noise power ratio coefficient. Or when the detection result of the received signal speech section detection unit is a non-speech section and the estimated value of the acoustic feedback gain is less than a predetermined threshold value continues for a predetermined time or more. It is characterized in that it is set to the transmission bias mode.
[0025]
The invention of claim 14 is claimed in claim 1 The line-side feedback gain for estimating the line-side feedback gain of the path that returns from the output point of the transmitting-side loss insertion means to the input point of the receiving-side loss insertion means via the far-end line echo path The deviation mode control means is characterized by not shifting to the reception deviation mode unless the estimated value of the line-side feedback gain satisfies a predetermined condition.
[0026]
The invention according to a fifteenth aspect is the invention according to the fourteenth aspect, wherein the bias mode control means is configured such that the estimated value of the near-end side noise power is an estimated value of the far-end side noise power and the first noise power ratio coefficient. When the detection result of the transmission signal voice section detection unit is a non-voice section and the estimated value of the line-side feedback gain is less than a predetermined threshold value continues for a predetermined time or more Is set to the reception bias mode.
[0027]
According to a sixteenth aspect of the present invention, in the twelfth aspect of the present invention, a line for a path that returns from the output point of the transmission side loss insertion means to the input point of the reception side loss insertion means via a line echo path on the far end side. Line-side feedback gain estimation means for estimating the side feedback gain, and the bias mode control means does not shift to the reception bias mode if the estimated value of the line-side feedback gain does not satisfy a predetermined condition. Each speech section detector on the receiver side includes a noise power estimator for estimating a noise power included in an input signal to the loss insertion means on the transmitter side or the receiver side, and the first or second instantaneous power estimate value. Loss insertion means for comparing a predetermined threshold and a product of the noise power estimated value and the predetermined value, respectively, and when the instantaneous power estimated value is larger than the threshold and the instantaneous power estimated value is larger than the product A speech / non-speech determination unit that determines that the input signal is a speech signal and determines that the input signal is a non-speech signal when the input signal is not large, and each speech section detection unit on the transmitting side and the receiving side includes the acoustic side and The threshold value and the predetermined value are changed according to the estimated value of each feedback gain estimating means on the line side.
[0028]
According to a seventeenth aspect of the present invention, in the twelfth aspect of the present invention, a line on a path for returning from the output point of the transmission side loss insertion means to the input point of the reception side loss insertion means via a line echo path on the far end side. Line-side feedback gain estimation means for estimating the side feedback gain, and the bias mode control means sets the reception bias mode when the estimated value of the line-side feedback gain satisfies a predetermined condition, Each speech section detector on the receiver side includes a noise power estimator for estimating a noise power included in an input signal to the loss insertion means on the transmitter side or the receiver side, and the first or second instantaneous power estimate value. Loss insertion means for comparing a predetermined threshold and a product of the noise power estimated value and the predetermined value, respectively, and when the instantaneous power estimated value is larger than the threshold and the instantaneous power estimated value is larger than the product What A voice / non-speech determination unit that determines an input signal as a speech signal and a non-speech signal when the input signal is not large, and the instantaneous power estimation unit is a filter having a characteristic that the rise is steep and the fall is gradual. The noise power estimation unit is composed of a filter having a characteristic that the rising edge is gradual and the falling edge is steep, and the speech section detection units on the transmitting side and the receiving side estimate the feedback gains on the acoustic side and the line side. The parameter for determining the characteristic is changed according to the estimated value of the means.
[0029]
In the invention of claim 18, in the invention of claim 17, the threshold value, the predetermined value, and the parameter can be individually set from the outside for the respective voice section detection units on the transmitting side and the receiving side. It is characterized by.
[0030]
The invention of claim 19 is characterized in that, in the invention of claim 17, the threshold value, the predetermined value, and the parameter can be set to different values depending on the type of the other call terminal.
[0031]
According to a twentieth aspect of the present invention, in the twelfth aspect of the present invention, a circuit for returning a path from the output point of the transmission side loss insertion means to the input point of the reception side loss insertion means via a line echo path on the far end side. Line-side feedback gain estimating means for estimating the side-side feedback gain, and the deviation mode control means does not shift to the reception-side deviation mode if the estimated value of the line-side feedback gain does not satisfy a predetermined condition, and the near-end side noise Referring to the difference between the estimated power value and the estimated far-end noise power, and the feedback gain estimated values on the acoustic side and the line side, the transmission bias mode setting amplification means and the reception bias mode setting amplification The gain of the means is adaptively updated.
[0032]
According to a twenty-first aspect of the present invention, in the twentieth aspect of the invention, an upper limit value is set for the gain of the transmission bias mode setting amplification unit and the reception bias mode setting amplification unit that are adaptively updated by the bias mode control unit. It is characterized by that.
[0033]
DETAILED DESCRIPTION OF THE INVENTION
(Embodiment 1)
In the present embodiment, since the basic configuration is the same as that of the conventional example, the same components are denoted by the same reference numerals and description thereof is omitted.
[0034]
In the present embodiment, as shown in FIG. 1, a first instantaneous power estimation unit 31 that estimates the instantaneous power of the input signal to the transmission side loss insertion means 1 and the instantaneous input signal to the reception side loss insertion means 2 Second instantaneous power estimation unit 32 for estimating power, and input to the transmission side loss insertion means 1 from the input point to the transmission side loss insertion means 1 and input to the reception side loss insertion means 2 through wraparound on the line side A line feedback gain multiplication unit 33 having a value determined according to the gain of the system that returns to the point as a coefficient, and a wraparound on the receiving side loss insertion unit 2 and the acoustic side from the input point to the receiving side loss insertion unit 2 The output signal of the acoustic feedback gain multiplication unit 34 having a value determined according to the gain of the path to the input point to the transmission side loss insertion means 1 and the second instantaneous power estimation unit 32 as an acoustic signal is transmitted. An output signal obtained by input to the feedback gain multiplier 34; The output obtained by inputting the output signal of the first instantaneous power estimation unit 31 to the line feedback gain multiplication unit 33 and the first comparator 35 that compares the magnitude relationship with the output signal of the first instantaneous power estimation unit 31 A second comparator 36 for comparing the magnitude relationship between the signal and the output signal of the second instantaneous power estimation unit 32, a transmission signal voice section detection unit 37 for detecting a voice section of the transmission signal, and a received signal A reception signal voice section detector 38 for detecting a voice section, a comparison result of the first comparator 35 and the second comparator 36, and a detection result of the transmission signal voice section detector 37 and the reception signal voice section detector 38. The insertion loss amount control means 3 includes an insertion loss amount distribution processing unit 30 that determines the call state based on the above and controls the insertion loss amounts of the transmission side loss insertion means 1 and the reception side loss insertion means 2. The mode control means 8 includes a far end side and a near end. The gains GT and GR of the transmission bias mode setting amplifier 6 and the reception bias mode setting amplifier 7 are substantially equal to each other based on the magnitude relationship of the estimated values of the noise powers and the detection result of the transmission signal voice section detection unit 37. The neutral mode is set to be equal (for example, 0 [dB]) and the gain GR of the reception bias mode setting amplifier 7 is set to any one of the reception bias modes in which the gain GT of the transmission bias mode setting amplifier 6 is increased. There is a feature in the point.
[0035]
In the first comparator 35, the output signal from the first instantaneous power estimator 31 and the output signal (first instantaneous power estimate) from the second instantaneous power estimator 32 are sent to the acoustic feedback gain multiplier 34. When the instantaneous power estimate is equal to or greater than the output signal of the acoustic feedback gain multiplier 34, the output signal C1 becomes 1, and the instantaneous power estimate is multiplied by the acoustic feedback gain. When it is less than the output signal of the means 34, the output signal C1 becomes zero. Further, in the second comparator 36, the output signal obtained by inputting the output signal of the first instantaneous power estimation unit 31 to the line feedback gain multiplication means 33 and the output signal of the second instantaneous power estimation unit 32 (first output) 2 when the output signal of the line side feedback gain multiplication means 33 is equal to or greater than the second instantaneous power estimation value, the output signal C2 becomes 1, and the line side feedback gain multiplication means 33 Is less than the second instantaneous power estimate, the output signal C2 becomes zero.
[0036]
The first and second instantaneous power estimators 31 and 32 are realized by an integration circuit or a digital filter having a characteristic that the rise is steep and the fall is gradual, and each of the input signals to the transmission side loss insertion means 1 And the instantaneous power of the signals amplified by the transmission bias mode setting amplifier 6 and the reception bias mode setting amplifier 7 respectively for the input signals to the reception side loss insertion means 2 are estimated.
[0037]
FIG. 2 is a block diagram showing a specific configuration of the transmission signal voice section detection unit 37 and the reception signal voice section detection unit 38. The transmission signal voice section detection unit 37 refers to a transmission signal (a reception signal in the reception signal voice section detection unit 38, and a parenthesis represents a case of the reception signal voice section detection unit 38). The background noise power estimation unit 37a (38a) for estimating the background noise level on the far end side), the instantaneous power estimation value Ps estimated by the first instantaneous power estimation unit 31 (32), and the background noise power estimation unit 37a ( Based on the background noise power estimation value Pn estimated in 38a), input signals (hereinafter collectively referred to as “reference signals”) in to the loss insertion means 1 and 2 on the transmission side or reception side are voiced. It is determined whether the signal is a signal or a non-speech signal. If it is determined to be a sound signal, the determination result (determination flag) SDF1 (SDF2) is set to 1. If it is determined to be a non-speech signal, the determination result SDF1 (SDF2) is determined. Together with the 0, the determination result SDF1 (sdf2) is and a voice / non-voice determining section 37b (38b) for holding the previous determination result SDF1 (sdf2) to update. Note that the background noise power estimation unit 37a (38a) is configured by an integration circuit or a digital filter having a characteristic that the rise is gradual and the fall is steep, and the background noise power estimation value is sequentially determined with reference to the reference signal in. Pn is updated, and the previous estimated value Pn is held until it is updated.
[0038]
On the other hand, the voice / non-voice determination unit 37b (38b), for example, compares the instantaneous power estimation value Ps output from the first instantaneous power estimation unit 31 (32) with a predetermined threshold value Ps0, and determines the instantaneous power estimation value. The ratio Ps / Pn between Ps and the background noise power estimation value Pn output from the background noise power estimation unit 37a (38a) is compared with a predetermined threshold value δ, and the instantaneous power estimation value Ps is compared with the threshold value Ps0. Is larger (Ps> Ps0) and the ratio Ps / Pn is larger than the threshold value δ (Ps / Pn> δ), it is determined as an audio signal and the determination result SDF1 (SDF2) is set to 1, and in other cases It is determined that the signal is a non-speech signal, and the determination result SDF1 (SDF2) is set to zero. Here, the threshold value Ps0 is a threshold value that defines the minimum level of the audio signal, and the threshold value δ is a threshold value that defines the minimum ratio between the audio signal level and the background noise level.
[0039]
In the insertion loss amount distribution processing unit 30 in the present embodiment, the comparison results C1 and C2 of the first and second comparators 35 and 36, and the detection results of the transmission signal voice section detection unit 37 and the reception signal voice section detection unit 38 The call state is determined with reference to SDF1 and SDF2, and the insertion loss amounts of the transmission side loss insertion means 1 and the reception side loss insertion means 2 are determined.
[0040]
Next, the operation of the bias mode control means 8 in this embodiment will be described with reference to the flowchart of FIG.
[0041]
First, the detection result (determination result) of the transmission signal voice section detector 37 is determined (step 1). If no voice section is detected (if SDF1 = 0), the neutral mode is changed to the reception bias mode. A transition determination is made (step 2), and it is determined whether the transition conditions are cleared (step 3). The transition condition here is the same as the transition condition from the neutral mode to the reception bias mode in the conventional example. If the transition condition is cleared, the gain GR of the reception bias mode setting amplifier 7 is set to G [dB], and the gain GT of the transmission bias mode setting amplifier 6 is set to 0 [dB]. (Step 6).
[0042]
On the other hand, when the speech section is detected by the transmission signal speech section detection unit 37 in Step 1 or when the transition condition is not cleared in Step 3, the transition determination from the neutral mode to the transmission bias mode is performed. (Step 4), and it is determined whether or not the transition condition is cleared (Step 5). The transition condition here is the same as the transition condition from the neutral mode to the transmission bias mode in the conventional example. If the transition condition is cleared, the gain GT of the transmission bias mode setting amplifier 6 is set to G [dB], and the gain GR of the reception bias mode setting amplifier 7 is set to 0 [dB]. If the mode is set (step 7) and the transition condition is not cleared, the gain GT of the transmission bias mode setting amplifier 6 and the gain GR of the reception bias mode setting amplifier 7 are both set to 0 [dB]. Set to neutral mode (step 8).
[0043]
Thus, in the present embodiment, when the speaker of the near-end terminal (speaking communication terminal equipped with the voice switching device of the present embodiment) is unilaterally speaking, the transmission signal voice section detection is performed. When the speech section is detected by the unit 37, the biasing mode control means 8 does not set the reception weighting mode regardless of the estimated values PFn and PNn of the two noise power estimation means 4 and 5, so the insertion loss amount control means 3 Is controlled so that the insertion loss amount by the transmission side loss insertion means 1 is maximized and the insertion loss amount by the reception side loss insertion means 2 is minimized and the call mode is set to the transmission mode. It is possible to prevent the insertion loss amount control means 3 from erroneously switching from the intermediate mode to the reception mode even when the speaker is speaking, and to prevent the voice from being interrupted or causing unnatural inflection. The intermediate mode is a state in which the insertion loss amounts of the loss insertion means 1 and 2 on the transmission side and the reception side are substantially the same, and the so-called idle mode (for example, the transmission side and the reception side are silent) Situation).
[0044]
(Embodiment 2)
The present embodiment is characterized by the bias mode control means 8, and since other configurations are the same as those of the first embodiment, illustration and description are omitted.
[0045]
As shown in FIG. 4, the bias mode control means 8 in the present embodiment includes a multiplier 82 for obtaining a product of an estimated value PFn of the far-end side noise power and a predetermined first noise power ratio coefficient X1, and this product The first comparator 81 that compares the estimated value PNn of the near-end side noise power, the inverter 83 that inverts the detection result SDF1 of the transmission signal voice section detector 37, the comparison result of the first comparator 81, and the inverter 83 An AND gate 84 for obtaining the logical product i of the detection result SDF1 inverted in step, a timing unit 85 including a counter that is incremented when the logical product i is 1 and reset when the logical product i is 0, and a timing unit 85 The comparison result Z of the second comparator 86 that compares the measured time duration (count value) T ′ with the first predetermined time T1 and the comparison result Z of the second comparator 86 Accordingly, there is provided a gain setting section 87 for setting the gain GT of the transmission bias mode setting amplifier 6 and the gain GR of the reception bias mode setting amplifier 7 to G [dB] or 0 [dB], respectively.
[0046]
The operation of the bias mode control means 8 will be specifically described. The product of the estimated value PFn of the far-end side noise power and the first noise power ratio coefficient X1 and the estimated value PNn of the near-end side noise power are the first comparator. A comparison result of 81 is output to the AND gate 84 when PFn × X1 <PNn and 0 when PFn × X1 ≧ PNn. The output (logical product) of the AND gate 84 is used only when the estimated value PNn of the near-end background noise power is larger than the product and no speech section is detected by the transmission signal speech section detection unit 37. ) I becomes 1, the duration T ′ in the time measuring means 85 is incremented, and the near-end side background noise power estimation value PNn is equal to or less than the product, or the speech signal speech segment detection unit 37 detects the speech segment If either condition is satisfied, the output i of the AND gate 84 becomes 0, and the duration T ′ in the time measuring means 85 is reset. The duration T ′ by the time measuring means 85 is compared with the first predetermined time T1 by the second comparator 86, and when the duration T ′ exceeds the first predetermined time T1, the second comparator 86 The output (comparison result) Z becomes 1, and the output Z becomes 0 if the duration T ′ does not exceed the first predetermined time T1. When the output Z of the second comparator 86 is 1, the gain setting unit 87 sets the gain GT of the transmission bias mode setting amplifier 6 to 0 [dB] and the gain GR of the reception bias mode setting amplifier 7 to G [ dB], the gain setting unit 87 sets the gain GT of the transmission bias mode setting amplifier 6 and the gain GR of the reception bias mode setting amplifier 7 when the output Z is 0. Is also set to 0 [dB] to set the neutral mode.
[0047]
As described above, in the bias mode control means 8 of the present embodiment, the near-end side noise power estimation value PNn is larger than the product of the far-end side noise power estimation value PFn and the first noise power ratio coefficient X1. In addition, the duration T ′ in a state in which no voice section is detected by the transmission signal voice section detector 37 is measured, and when the duration T ′ becomes equal to or longer than the first predetermined time T1, the amplification means 7 for setting the reception bias mode. In order to set the reception weighted mode as G [dB] and the gain GT of the transmission weighting mode setting amplification means 6 as 0 [dB], the first noise power ratio coefficient X1 and the first predetermined time T1 are set. By setting the value to an appropriate value, it is possible to adjust the ease of transition from the neutral mode to the reception weighted mode by the weighted mode control means 8, and the surroundings of the voice call terminal equipped with the voice switching device of the present embodiment To noise It is possible to arbitrarily set the Switching Characteristics of.
[0048]
(Embodiment 3)
Since the basic configuration of this embodiment is the same as that of the conventional example and Embodiment 1 as shown in FIG. 5, the same components are denoted by the same reference numerals, and the description thereof is omitted. Only the configuration will be described. In the present embodiment, the deviation mode control means 8 transmits the transmission deviation based on the magnitude relationship between the noise power estimation values PFn and PNn on the far end side and the near end side and the detection result SDF2 of the reception signal speech section detection unit 38. It is characterized in that it is set to either mode or neutral mode.
[0049]
Next, the operation of the bias mode control means 8 in this embodiment will be described with reference to the flowchart of FIG.
[0050]
First, the detection result (determination result) of the reception signal voice section detector 38 is determined (step 1). If no voice section is detected (if SDF2 = 0), the neutral mode is changed to the transmission bias mode. A transition determination is made (step 2), and it is determined whether the transition conditions are cleared (step 3). The transition condition here is the same as the transition condition from the neutral mode to the transmission bias mode in the conventional example. If the transition condition is cleared, the gain GR of the reception bias mode setting amplifier 7 is set to 0 [dB], and the gain GT of the transmission bias mode setting amplifier 6 is set to G [dB]. The mode is set (step 6).
[0051]
On the other hand, when a speech section is detected by the received signal speech section detection unit 38 in step 1, or when the transition condition is not cleared in step 3, the transition determination from the neutral mode to the reception bias mode is performed ( Step 4), it is determined whether or not the transition condition is cleared (Step 5). The transition condition here is the same as the transition condition from the neutral mode to the reception bias mode in the conventional example. If the transition condition is cleared, the gain GT of the transmission bias mode setting amplifier 6 is set to 0 [dB], and the gain GR of the reception bias mode setting amplifier 7 is set to G [dB]. If the transition condition is not cleared, the gain GT of the transmission bias mode setting amplifier 6 and the gain GR of the reception bias mode setting amplifier 7 are both set to 0 [dB]. The mode is set (step 8).
[0052]
Thus, in this embodiment, the speaker of the far-end side terminal (a voice call terminal equipped with the voice switching device of this embodiment and another call terminal constituting the voice call system) speaks unilaterally. In this case, when the speech signal is detected by the reception signal speech section detector 38, the bias mode control means 8 performs the transmission bias mode regardless of the estimated values PFn and PNn of the two noise power estimation means 4 and 5. Since the insertion loss amount control means 3 is controlled so that the insertion loss amount by the transmission side loss insertion means 1 is minimized and the insertion loss amount by the reception side loss insertion means 2 is maximized, the call mode is set to the reception mode. Therefore, it is possible to prevent the insertion loss amount control means 3 from erroneously switching from the neutral mode to the transmission mode even when the far-end speaker is speaking, and the voice is interrupted or unnatural. It is possible to suppress the intonation from occurring.
[0053]
As in the first embodiment, the bias mode control means 8 uses the magnitude relationship between the estimated values PFn and PNn of the noise power on the far end side and the near end side and the detection result SDF1 of the transmission signal speech section detection unit 37. On the basis of this, if either the reception bias mode or the neutral mode is set, the insertion loss amount control means 3 is prevented from erroneously switching from the neutral mode to the transmission mode or the reception mode, and the sound is interrupted or not generated. Natural inflection can be suppressed more reliably.
[0054]
(Embodiment 4)
The present embodiment is characterized by the bias mode control means 8, and since other configurations are the same as those of the third embodiment, illustration and description thereof are omitted.
[0055]
As shown in FIG. 7, the bias mode control means 8 in the present embodiment includes a multiplier 82 ′ for obtaining a product of the near-end side noise power estimated value PNn and a predetermined second noise power ratio coefficient X2, and this product. Comparison result between the third comparator 81 ′ that compares the estimated value PFn with the far-end noise power, the inverter 83 ′ that inverts the detection result SDF2 of the reception signal speech section detector 38, and the third comparator 81 ′. AND gate 84 ′ for obtaining the logical product i ′ of the detection result SDF 2 inverted by the inverter 83 ′, and a timing means comprising a counter that is incremented when the logical product i ′ is 1 and reset when it is 0 85 ′, a fourth comparator 86 ′ that compares the time (count value) T ″ measured by the time measuring means 85 ′ and the second predetermined time T2, and a comparison result Z ′ of the fourth comparator 86 ′. And a gain setting unit 87 ′ for setting the gain GT of the transmission bias mode setting amplifier 6 and the gain GR of the reception bias mode setting amplifier 7 to G [dB] or 0 [dB], respectively.
[0056]
The operation of the bias mode control means 8 will be specifically described. The product of the near-end side noise power estimated value PNn and the second noise power ratio coefficient X2 and the far-end side noise power estimated value PFn are the third comparator. A comparison result of 81 'is output to the AND gate 84' when PNn × X2 <PFn, and when PNn × X2 ≧ PFn, 0 is output. The output (logical product) of the AND gate 84 ′ is obtained only when the estimated value PFn of the far-end background noise power is larger than the product and no speech section is detected by the received signal speech section detection unit 38. ) I ′ becomes 1, and the duration T ″ in the time measuring means 85 ′ is incremented, and the estimated value PFn of the far-end background noise power does not become larger than the product, or the speech signal speech interval detection unit 38 determines the speech interval Or one of the conditions is met, the output i ′ of the AND gate 84 ′ becomes 0, and the duration T ″ in the time measuring means 85 ′ is reset. The duration T ″ by the time measuring means 85 ′ is compared with the second predetermined time T2 by the fourth comparator 86 ′. When the duration T ″ exceeds the second predetermined time T2, the fourth comparator If the output (comparison result) Z ′ of 86 ′ becomes 1 and the duration T ″ does not exceed the second predetermined time T2, the output Z ′ becomes 0. Then, the output Z ′ of the fourth comparator 86 ′ In the case of 1, the gain setting unit 87 ′ sets the gain GT of the transmission bias mode setting amplifier 6 to G [dB] and the gain GR of the reception bias mode setting amplifier 7 to 0 [dB]. When the output Z ′ is set to 0, the gain setting unit 87 ′ sets both the gain GT of the transmission bias mode setting amplifier 6 and the gain GR of the reception bias mode setting amplifier 7 to 0 [dB]. By doing so, the neutral mode is set.
[0057]
Thus, in the bias mode control means 8 of the present embodiment, the far-end side noise power estimate value PFn is larger than the product of the near-end side noise power estimate value PNn and the second noise power ratio coefficient X2. In addition, the duration T ″ of the state in which no speech segment is detected by the received signal speech segment detection unit 38 is timed, and when the duration T ″ is equal to or longer than the second predetermined time T2, the amplification unit 7 for setting the reception bias mode is set. Since the gain GR is set to 0 [dB] and the gain GT of the transmission bias mode setting amplifier 6 is set to G [dB] to set the transmission bias mode, the second noise power ratio coefficient X2 and the second predetermined time T2 are set. Is set to an appropriate value, the ease of transition from the neutral mode to the transmission bias mode by the bias mode control means 8 can be adjusted, and the voice call terminal equipped with the voice switching device of this embodiment can be adjusted. Ambient noise It can be set arbitrarily Switching Characteristics against.
[0058]
(Embodiment 5)
By the way, in the voice switching devices of the above-described first to fourth embodiments, it is of course possible to configure each means by hardware, but a single processor such as a digital signal processor (DSP) is used, and the DSP It is desirable to realize each of the above-mentioned means by controlling the hardware of the above with dedicated software. The voice switching device VS of this embodiment is realized by a combination of a DSP and dedicated software.
[0059]
FIG. 8 shows a schematic configuration of a loudspeaker terminal equipped with the voice switching device VS of the present embodiment. The DSP 100 has a series of call processing functions and a function for generating a notification sound, an alarm sound, or an alarm sound. The voice switching device VS is configured as a part of the function of the DSP 100. Further, reference numeral 200 in the figure denotes a CPU, which reads, for example, dedicated software stored in the ROM 201 into the memory, for example, a function for detecting a call from the other party's call terminal and causing the DSP 100 to start call processing. It is realized by executing. The DSP 100 and the CPU 200 are connected via a communication interface such as a serial port or a parallel port.
[0060]
In the loudspeaker terminal according to the present embodiment, the threshold value Ps0 used to detect the voice section in the transmission signal voice section detection unit 37 or the reception signal voice section detection unit 38 from the CPU 200 to the DSP 100 when starting a call. The data of δ is transmitted from the CPU 200 to the DSP 100 via the communication interface, and this is initialized with respect to the software module that realizes the voice section detection units 37 and 38 by the threshold setting means 101 provided in the DSP 100. After the values (data) of the threshold values Ps0 and δ are initialized, the DSP 100 activates the voice switching device VS, the echo canceller, and the like to start the call processing, and the CPU 200 controls the hardware of the voice call terminal. To establish a communication path with the other party's telephone terminal. The threshold value setting means 101 is also realized by software.
[0061]
The present embodiment is configured as described above, and the threshold values Ps0 and δ used for detection of a voice section in the transmission signal voice section detection unit 37 or the reception signal voice section detection unit 38 are externally (CPU 200). Since it is possible to set, the versatility of the voice switching device VS increases, and it can be easily adapted to various voice call terminals.
[0062]
(Embodiment 6)
FIG. 9 shows a schematic configuration of a loudspeaker call terminal provided with the voice switching device VS of the present embodiment, and the same reference numerals are given to components common to the fifth embodiment.
[0063]
In the loudspeaker terminal according to the present embodiment, when starting a call, the CPU 200 instructs the DSP 100 to use parameters for detecting a voice section in the transmission signal voice section detection unit 37 or the reception signal voice section detection unit 38, specifically, Transmits data of time constants (parameters for determining the rising and falling characteristics of the digital filter) used in the calculation of the instantaneous power estimation unit 91 and the background noise power estimation unit 92 from the CPU 200 to the DSP 100 via the communication interface. Then, the time constant setting means 102 provided in the DSP 100 initializes the software module that implements the voice section detection units 37 and 38. After these time constants are initialized, the DSP 100 activates the voice switching device VS, the echo canceller, and the like to start the call processing, and the CPU 200 controls the hardware of the loudspeaker call terminal to communicate with the other call terminal. Form a call between. The time constant setting means 102 is also realized by controlling the hardware of the DSP 100 with software.
[0064]
This embodiment is configured as described above, and the time constant used for detection of the voice section in the transmission signal voice section detection unit 37 or the reception signal voice section detection unit 38 can be set from the outside (CPU 200). As a result, the versatility of the voice switching device VS increases, and it can be easily adapted to various loudspeaker terminals.
[0065]
(Embodiment 7)
FIG. 10 shows a schematic configuration of a loudspeaker call terminal provided with the voice switching device VS of the present embodiment, and the same reference numerals are given to components common to the fifth embodiment.
[0066]
In the loudspeaker terminal according to the present embodiment, threshold values Ps0 and δ (referred to as “transmission side threshold values”) and a time constant (which are referred to as “transmission side time constants) for the transmission signal voice section detection unit 37. ) And threshold values Ps0 and δ (referred to as “received-side threshold values”) and time constants (referred to as “received-side time constants”) for the received signal voice section detector 38. It is stored in the flash memory 110.
[0067]
On the other hand, the DSP 100 receives a command from the CPU 200 via the communication interface, interprets the command and performs necessary processing, and transmits and receives each command to the voice switching device VS. Parameter setting means 104 is provided for setting a threshold and a time constant for a software module that implements the speech section detection units 37 and 38. However, the command processing means 103 and the parameter setting means 104 can also be realized by software.
[0068]
When a call is started, first, a call start request command is transmitted from the CPU 200 to the DSP 100 via the communication interface, and the command processing means 103 that has received the call start request command interprets the content and transmits the call. Address data indicating the storage location of the parameters used in the signal voice section detector 37 and the reception signal voice section detector 38 is given to the parameter setting means 104. Based on the received address data, the parameter setting means 104 reads the transmission side and reception side threshold values and time constant data from the flash memory 110, and sets the initial values for the software modules that implement the speech section detection units 37 and 38. Set. After these initial settings are completed, the parameter setting means 104 sets a flag indicating that the parameter initialization is completed, and the command processing means 103 transmits a response command to the CPU 200 in response to the setting of the flag.
[0069]
Thus, the transfer characteristic from the microphone (not shown) of the near-end side terminal (speaking telephone terminal in the present embodiment) to the reference point of the transmission signal in the voice switching device VS and the microphone (not shown) of the far-end side terminal. The transmission characteristics from the reference signal to the reference point of the received signal are generally different, but in this embodiment, threshold values Ps0 and δ and time constants for the transmission signal voice section detector 37 and the received signal transmission voice section detector 10 are set. Since each can be individually set from the outside (CPU 200), the difference between the two transfer characteristics as described above can be easily corrected.
[0070]
(Embodiment 8)
FIG. 11 shows a schematic configuration of a loudspeaker call terminal provided with the voice switching device VS of the present embodiment, and the same reference numerals are given to components common to the seventh embodiment.
[0071]
As a call system in which the voice switching device of the present invention is used, a lobby intercom installed in an entrance lobby in an apartment house such as a condominium, a doorphone installed in an entrance of each dwelling unit, and a parent unit installed in each dwelling unit There are those that make loud voice calls between each other. In such a call system, the transfer characteristic up to the reference point of the received signal described in the seventh embodiment differs depending on the partner terminal. That is, when the loudspeaker terminal equipped with the voice switching device VS is the master unit, the transmission characteristic of the received signal in the call with the door phone and the transmission characteristic of the received signal in the call with the lobby interphone are the line length of the call, etc. Therefore, the difference between the transmission characteristic of the transmission signal and the transmission characteristic of the reception signal is naturally different for each call terminal on the other side.
[0072]
Therefore, in the present embodiment, different values can be set according to the type of the other party's call terminal (in the above example, door phone and lobby interphone). Specifically, the two areas M1, M2 stores doorphone parameters (transmission side and reception side thresholds and time constants) and lobby intercom parameters, respectively. Then, the CPU 200 designates the address of the parameter according to the type of the call terminal of the call partner with a command, and based on the address data received by the parameter setting means 104, the transmitter and the call receiver according to the type of the call terminal of the other party Side threshold value and time constant data are selected and read from the areas M1 and M2 of the flash memory 110, and are initialized for the software modules that implement the speech section detection units 37 and 38.
[0073]
Thus, in this embodiment, the threshold values Ps0 and δ and the time constant for the transmission signal voice section detection unit 37 and the reception signal voice section detection unit 38 are individually determined from the outside (CPU 200) for each call terminal on the other side. Therefore, it is possible to easily correct the difference between two transfer characteristics that differ for each call terminal on the other side.
[0074]
(Embodiment 9)
A block diagram of this embodiment is shown in FIG. The difference between this embodiment and Embodiment 1 and Embodiment 3 is that the near-end side noise power estimation means 4 is also used by the background noise power estimation unit 37a (see FIG. 2) that constitutes the transmission side speech section detection unit 37. At the same time, the far-end side noise power estimation means 5 is shared by the background noise power estimation unit 38a (see FIG. 2) constituting the reception side voice section detection unit 38. Since other configurations and operations are the same as those in the first or third embodiment, the description thereof is omitted.
[0075]
That is, in the present embodiment, the configuration can be simplified as compared to the first or third embodiment, and the amount of calculation in software can be reduced when the DSP 100 and software are used.
[0076]
(Embodiment 10)
The present embodiment is characterized by the bias mode control means 8, and since other configurations are the same as those of the first embodiment, illustration and description are omitted.
[0077]
The basic configuration of the bias mode control means 8 in this embodiment is the same as that in the second embodiment. As shown in FIG. 13, the first comparator 81, the multiplier 82, the inverter 83, the AND gate 84, and the time measuring means 85. In addition to the second comparator 86 and the gain setting unit 87, a multiplier 182 for obtaining a product of the estimated value PFn of the far-end side noise power and a predetermined third noise power ratio coefficient X3, and this product and the near end A fifth comparator 181 that compares the estimated value PNn of the side noise power, a time counting means 183 that includes a counter that is incremented when the output of the fifth comparator 181 is 1 and reset when it is 0, Time measured by means 183 (count value) T 2 And a third comparator 184 that compares the third predetermined time T3, and an OR gate 88 that obtains the logical sum Z of the comparison result Y1 of the second comparator 86 and the comparison result Y2 of the sixth comparator 184. The third noise power ratio coefficient X3 is set to a value that is several times greater than the first noise power ratio coefficient X1, and the third predetermined time T3 is set to a value that is greater than the first predetermined time T1. .
[0078]
Next, the operation of the bias mode control means 8 in this embodiment will be described. However, the description of the operation of the same configuration as the second embodiment is omitted, and only the operation of the configuration added in the present embodiment to the second embodiment and the entire operation will be described.
[0079]
The product of the estimated value PFn of the far-end side noise power and the third noise power ratio coefficient X3 and the estimated value PNn of the near-end side noise power are compared in the fifth comparator 181, and 1 is obtained when PFn × X3 <PNn. , PFn × X3 ≧ PNn, a comparison result of 0 is output to the time measuring means 183. When the estimated value PNn of the near-end side background noise power is larger than the product, the duration T in the time measuring means 183 is determined. 2 'Is incremented and the estimated value PNn of the near-end side background noise power does not become larger than the product, or the voice signal is detected by the transmission signal voice signal detection unit 37, and the time is measured when any of the conditions is satisfied. Duration T in means 183 2 'Is reset. Duration T by time measuring means 183 2 'Is compared with the third predetermined time T3 by the sixth comparator 184, and the duration T 2 When 'exceeds the third predetermined time T3, the output (comparison result) Y2 of the sixth comparator 184 becomes 1, and the duration T 2 If 'does not exceed the third predetermined time T3, the output Y2 becomes zero.
[0080]
The output Y2 of the sixth comparator 184 and the output Y1 of the second comparator 86 are inputted to the OR gate 88. If at least one of the outputs Y1 and Y2 is 1, the output Z of the OR gate 88 becomes 1. If both outputs Y1 and Y2 are 0, the output Z of the OR gate 88 is also 0. Further, when the output Z of the OR gate 88 is 1, the gain setting unit 87 sets the gain GT of the transmission bias mode setting amplifier 6 to 0 [dB] and the gain GR of the reception bias mode setting amplifier 7 to G [dB]. By setting, the reception bias mode is set, and when the output Z is 0, the gain setting unit 87 sets both the gain GT of the transmission bias mode setting amplifier 6 and the gain GR of the reception bias mode setting amplifier 7 to 0 [ dB] is set to the neutral mode.
[0081]
Thus, in the second embodiment, when there is a high level of unsteady noise (for example, broadcast sound of television or radio) around the near-end call terminal, the transmission signal voice section detection unit 37 is If the voice section is erroneously detected, it is difficult to set the reception bias mode, and there is a possibility that the call mode is fixed to the transmission mode (so-called one-sided fall). On the other hand, the bias mode control means 8 in the present embodiment uses the estimated value PNn of the near-end side noise power as the estimated value of the far-end side noise power even if the duration T ′ is less than the first predetermined time T1. When the state larger than the product of PFn and the third noise power ratio coefficient X3 continues for the third predetermined time T3 or longer, the reception biased mode is set. Even in the presence of a high level of unsteady noise, it is difficult to set the reception bias mode and it is possible to prevent the occurrence of one-sided fall.
[0082]
(Embodiment 11)
The present embodiment is characterized by the bias mode control means 8, and since other configurations are the same as those of the third embodiment, illustration and description thereof are omitted.
[0083]
The bias mode control means 8 in this embodiment has the same basic configuration as that of the fourth embodiment, and as shown in FIG. 14, a third comparator 181 ′, a multiplier 82 ′, an inverter 83 ′, and an AND gate 84 ′. In addition to the time measuring means 85 ′, the fourth comparator 86 ′, and the gain setting unit 87 ′, a multiplier 182 that obtains a product of the estimated value PNn of the near-end side noise power and a predetermined fourth noise power ratio coefficient X4. 'And a seventh comparator 181' that compares this product with the estimated value PFn of the far-end noise power, and incremented when the output of the seventh comparator 181 'is 1, and reset when it is 0. A time measuring means 183 'comprising a counter and a time measured by the time measuring means 183' (count value) T 2 ”And the fourth predetermined time T4, or OR gate 88 ′ for obtaining the logical sum Z ′ of the comparison result Y3 of the fourth comparator 86 ′ and the comparison result Y4 of the eighth comparator 184 ′. The fourth noise power ratio coefficient X4 is set to a value that is several times the second noise power ratio coefficient X2, and the fourth predetermined time T4 is longer than the second predetermined time T2. Set to a value.
[0084]
Next, the operation of the bias mode control means 8 in this embodiment will be described. However, the description of the operation of the same configuration as the fourth embodiment is omitted, and only the operation of the configuration added in the present embodiment to the fourth embodiment and the entire operation will be described.
[0085]
The product of the estimated value PNn of the near-end side noise power and the fourth noise power ratio coefficient X4 and the estimated value PFn of the far-end side noise power are compared in the seventh comparator 181 ′, and when PNn × X4 <PFn 1. When PNn × X4 ≧ PFn, a comparison result of 0 is output to the timing means 183 ′. When the estimated value PFn of the far-end side background noise power is larger than the product, the duration T in the time measuring means 183 ′ 2 "Is incremented and the far end side background noise power estimated value PFn does not become larger than the product, or the voice signal is detected by the reception signal voice period detector 38, the time measuring means is satisfied. Duration T at 183 ' 2 "Is reset. Duration T by time measuring means 183 ' 2 "Is compared with the fourth predetermined time T4 by the eighth comparator 184 ', and the duration T 2 ”Exceeds the fourth predetermined time T4, the output (comparison result) Y4 of the eighth comparator 184 ′ becomes 1, and the duration T 2 If "" does not exceed the fourth predetermined time T4, the output Y4 becomes zero.
[0086]
The output Y4 of the eighth comparator 184 ′ and the output Y3 of the fourth comparator 86 ′ are input to the OR gate 88 ′. If at least one of the outputs Y3 and Y4 is 1, the output of the OR gate 88 ′ is output. If Z ′ is 1 and both outputs Y3 and Y4 are 0, the output Z ′ of the OR gate 88 ′ is also 0. Further, when the output Z ′ of the OR gate 88 ′ is 1, the gain setting unit 87 ′ sets the gain GT of the transmission bias mode setting amplifier 6 to G [dB], and the gain GR of the reception bias mode setting amplifier 7 to 0 [ dB] is set to the transmission bias mode, and when the output Z ′ is 0, the gain setting unit 87 ′ has the gain GT of the transmission bias mode setting amplifier 6 and the gain of the reception bias mode setting amplifier 7. The neutral mode is set by setting GR to 0 [dB].
[0087]
Thus, in the fourth embodiment, when there is a high level of unsteady noise (for example, wind noise or noise caused by construction) around the far-end telephone terminal, the received signal voice section detecting unit 38 is used. However, if the voice section is erroneously detected, it is difficult to set the transmission eccentric mode, and there is a possibility that the call mode is fixed to the reception mode (so-called one-sided fall). On the other hand, the bias mode control means 8 in the present embodiment uses the estimated value PFn of the far-end side noise power as the estimated value of the near-end side noise power even if the duration T ″ is less than the second predetermined time T2. When the state larger than the product of PNn and the fourth noise power ratio coefficient X4 continues for the fourth predetermined time T4 or longer, the mode shifts to the transmission bias mode. Even in the presence of a high level of unsteady noise, the transition to the transmission bias mode is reliably performed, so that the occurrence of one-sided fall can be prevented.
[0088]
Embodiment 12
As shown in FIG. 15, the voice switching device according to the present embodiment has an acoustic echo path H on the near end side from the output point of the receiving side loss insertion means 2 in addition to the configuration of the ninth embodiment. AC The acoustic side feedback gain estimation means 11 for estimating the acoustic side feedback gain α of the path returning to the input point of the transmission side loss insertion means 1 is provided, and the estimated value | α ′ | If the above condition is not satisfied, the deviation mode control means 8 does not shift to the reception deviation mode.
[0089]
The acoustic side feedback gain estimation means 11 estimates the time-average power of the input signal (transmission signal) of the transmission side loss insertion means 1 in a short time, and the input signal (reception signal) of the reception side loss insertion means 2. Estimate the time average power in a short time, and further return the acoustic side return path H AC The minimum value of the estimated value of the time average power of the output signal of the receiver side loss insertion means 2 is obtained at the maximum delay time assumed in FIG. 1, and the time average power of the input signal of the transmitter side loss insertion means 1 is calculated with this minimum value. A value obtained by dividing the estimated value is an estimated value | α ′ | of the acoustic feedback gain α.
[0090]
In each of the above embodiments, since the gain of the transmission bias mode setting amplifier 6 is increased in the transmission bias mode, so-called reception blocking is likely to occur. Here, reception blocking refers to the comparison between the first and second acoustic echo signals generated by acoustic coupling between a speaker and a microphone on the near end when sound is input from the far end while the near end is silent. When the comparison result of either or both of the devices 35 and 36 is inverted from the reception state to the transmission state, the call mode estimated in the insertion loss amount distribution processing unit 30 is not the reception mode, and the reception side loss insertion means 2 This is a phenomenon in which an unnatural break occurs when listening to the voice input from the far end side at the near end side, or when the amount of loss is large, the loss is inserted. On the other hand, in this embodiment, if the state where the estimated value | α ′ | of the acoustic feedback gain α is equal to or less than a predetermined threshold value α0 does not continue for a certain time or longer, the eccentric mode control means 8 performs the transmission eccentric mode. Therefore, incoming call blocking can be prevented.
[0091]
Note that the method for obtaining the estimated value | α ′ | of the acoustic feedback gain α is not limited to the above method, and a conventionally known method described in the specification of a patent application already filed by the present applicant is adopted. It doesn't matter.
[0092]
(Embodiment 13)
The present embodiment is characterized by the bias mode control means 8, and the other configurations are the same as those of the twelfth embodiment, so illustration and description thereof are omitted.
[0093]
In the present embodiment, the bias mode control means 8 has an estimated value PFn of the far-end side noise power equal to or greater than a product of the estimated value PNn of the near-end side noise power and the second noise power ratio coefficient X2, and the received signal If the detection result SDF2 of the speech section detecting means 10 is not a speech section and the estimated value | α '| of the acoustic side feedback gain α is not more than a predetermined threshold value α0 does not continue for a predetermined time or more, It is characterized in that the occurrence of reception blocking is suppressed by not shifting to the mode.
[0094]
The bias mode control means 8 in this embodiment has the same basic configuration as that of the fourth embodiment, and as shown in FIG. 16, a third comparator 81 ′, a multiplier 82 ′, an inverter 83 ′, and an AND gate 84 ′. A tenth comparator 185 ′ that compares the estimated value | α ′ | of the acoustic feedback gain α with a threshold value α0 in addition to the time measuring means 85 ′, the fourth comparator 86 ′, and the gain setting unit 87 ′; A second AND gate 186 ′ for obtaining a logical product of the output of the AND gate 84 ′ and the comparison result of the tenth comparator 185 ′, and the output of the second AND gate 186 ′ is used as an input of the timing means 85 ′. Yes.
[0095]
Next, the operation of the bias mode control means 8 in this embodiment will be described. However, the description of the operation of the same configuration as the fourth embodiment is omitted, and only the operation of the configuration added in the present embodiment to the fourth embodiment and the entire operation will be described.
[0096]
The estimated value | α ′ | of the acoustic side feedback gain α and the threshold value α0 are compared by the tenth comparator 185 ′, and when the estimated value | α ′ | is less than the threshold value α0, the estimated value | α ′. When | is equal to or greater than the threshold value α0, a comparison result of 0 is output to the second AND gate 186 ′. Therefore, the estimated value PFn of the far-end side background noise power is a value larger than the product, the speech section is not detected by the reception signal speech section detection unit 38, and the estimated value | α of the acoustic side feedback gain α Only when “|” is less than the threshold value α0, the duration T ”in the time measuring means 85 ′ is incremented and the estimated value PFn of the far-end background noise power does not become larger than the product or the received signal speech section detecting unit If the condition of either the voice section is detected by 38 or the estimated value | α ′ | of the acoustic feedback gain α is equal to or greater than the threshold value α0, the duration T ″ in the time measuring means 85 ′ is set. Reset. Then, when the duration T ″ by the time measuring means 85 ′ exceeds the second predetermined time T2, the output Z ′ of the fourth comparator 86 ′ becomes 1, and the gain setting unit 87 ′ is used for setting the transmission bias mode. By setting the gain GT of the amplifier 6 to G [dB] and the gain GR of the amplifier 7 for setting the reception bias mode to 0 [dB], the transmission bias mode is set, and the duration T ″ is the second predetermined time T2. If it does not exceed, the output Z ′ becomes 0, and the gain setting unit 87 ′ sets both the gain GT of the transmission bias mode setting amplifier 6 and the gain GR of the reception bias mode setting amplifier 7 to 0 [dB]. To set the neutral mode.
[0097]
Thus, when the estimated value | α ′ | of the acoustic feedback gain α is equal to or greater than the threshold value α0, the bias mode control means 8 is always set to the neutral mode regardless of other conditions, and the acoustic feedback gain α is Since the transmission bias mode is not set in a relatively large situation, occurrence of reception blocking can be suppressed. Furthermore, since the threshold value α0 can be used as a variable parameter, it is possible to set various device performances according to the installation environment of the device.
[0098]
(Embodiment 14)
As shown in FIG. 17, the voice switching apparatus according to the present embodiment has a line echo path H on the far end side from the output point of the transmission side loss insertion means 1 in addition to the configuration of the ninth embodiment. LIN Line-side feedback gain estimation means 12 for estimating the line-side feedback gain β of the path to be fed back to the input point of the receiving-side loss insertion means 2 via the communication line, and the estimated value | β ′ | If the condition is not satisfied, the feature is that the deviation mode control means 8 does not shift to the transmission deviation mode.
[0099]
The line-side feedback gain estimating means 12 estimates the time average power of the input signal (received signal) of the receiving side loss inserting means 2 in a short time and the input signal (transmitted signal) of the transmitting side loss inserting means 1. Estimate the time average power in a short time, and then the line side feedback path H LIN The minimum value of the estimated value of the time average power of the output signal of the transmission side loss insertion means 1 is obtained at the maximum delay time assumed in FIG. 1, and the time average power of the input signal of the reception side loss insertion means 2 is calculated with this minimum value. A value obtained by dividing the estimated value is an estimated value | β ′ | of the line-side feedback gain β.
[0100]
In each of the embodiments described above, so-called transmission blocking is likely to occur because the gain of the reception bias mode setting amplifier 7 is increased in the reception bias mode. Here, transmission blocking is a line echo signal generated by acoustic coupling on the far end side or signal wraparound in the 2-wire to 4-wire conversion circuit when speech is input from the near end side with the far end side being silent. As a result of the comparison of either one or both of the first and second comparators 35 and 36 from the transmission state to the reception state, the call mode estimated by the insertion loss amount distribution processing unit 30 is changed to the transmission mode. Since a loss is inserted into the transmission side loss insertion means 1, an unnatural interruption occurs when listening to the voice input from the near end side at the far end side, or the loss amount is large. A phenomenon that cannot be heard at all. On the other hand, in this embodiment, unless the state where the estimated value | β ′ | of the line-side feedback gain β is equal to or less than the predetermined value β0 continues for a certain time or longer, the bias mode control means 8 does not shift to the transmission bias mode. Therefore, transmission blocking can be prevented.
[0101]
Note that the method of obtaining the estimated value | β ′ | of the line side feedback gain β is not limited to the above method, and a conventionally known method described in the specification of the patent application already filed by the present applicant is adopted. It doesn't matter.
[0102]
(Embodiment 15)
The present embodiment is characterized by the bias mode control means 8, and the other configurations are the same as those of the thirteenth embodiment, so illustration and description thereof are omitted.
[0103]
In the present embodiment, the bias mode control means 8 has the estimated value PNn of the near-end side noise power equal to or greater than the product of the estimated value PFn of the far-end side noise power and the first noise power ratio coefficient X1. When the detection result SDF1 of the signal voice section detector 37 is not a voice section and the estimated value | β '| of the line-side feedback gain β is equal to or less than a predetermined threshold value β0 continues for a predetermined time or more The mode is characterized in that transmission blocking is suppressed by setting the mode.
[0104]
The bias mode control means 8 in this embodiment has the same basic configuration as that of the second embodiment, and as shown in FIG. 18, a first comparator 81, a multiplier 82, an inverter 83, an AND gate 84, and a time measuring means 85. In addition to the second comparator 86 and the gain setting unit 87, the ninth comparator 185 that compares the estimated value | β ′ | of the line-side feedback gain β with the threshold value β0, the output of the AND gate 84, and the ninth And a second AND gate 186 for obtaining the logical product of the comparison results of the comparator 185, and the output of the second AND gate 186 is used as the input of the time measuring means 85.
[0105]
Next, the operation of the bias mode control means 8 in this embodiment will be described. However, the description of the operation of the same configuration as the second embodiment is omitted, and only the operation of the configuration added in the present embodiment to the second embodiment and the entire operation will be described.
[0106]
The estimated value | β ′ | of the line-side feedback gain β and the threshold value β0 are compared by the ninth comparator 185. When the estimated value | β ′ | is less than the threshold value β0, the estimated value | β ′ | Is equal to or greater than the threshold value β0, the comparison result of 0 is output to the second AND gate 186. Accordingly, the estimated value PNn of the near-end side background noise power is a value larger than the product, the speech section is not detected by the transmission signal speech section detection unit 37, and the estimated value of the line-side feedback gain β | Only when β ′ | is less than the threshold value β0, the duration T ′ in the time measuring means 85 is incremented, and the estimated value PNn of the near-end side background noise power is equal to or less than the product, or the transmission signal speech section detection unit 37, the duration T ′ in the time measuring means 85 is reset when the condition of either the voice section is detected or the estimated value | β ′ | of the line side feedback gain β is equal to or greater than the threshold value β0. Is done. Then, when the duration T ′ by the time measuring means 85 exceeds the first predetermined time T1, the output Z of the second comparator 86 becomes 1, and the gain setting unit 87 obtains the gain of the transmission bias mode setting amplifier 6. By setting GT to 0 [dB] and the gain GR of the reception bias mode setting amplifier 7 to G [dB], the reception bias mode is set. If the duration T ′ does not exceed the first predetermined time T1, the output is performed. Z is set to 0, and the gain setting unit 87 sets the gain GT of the transmission bias mode setting amplifier 6 and the gain GR of the reception bias mode setting amplifier 7 to 0 [dB], so that the neutral mode is set. The
[0107]
Thus, when the estimated value | β ′ | of the line side feedback gain β is equal to or greater than the threshold value β0, the bias mode control means 8 is always set to the neutral mode regardless of other conditions, and the line side feedback gain β is In a relatively large situation, the incoming call bias mode is not set, so that transmission blocking can be suppressed. Further, since the threshold value β0 can be used as a variable parameter, the performance of the device can be set variously according to the installation environment of the device.
[0108]
(Embodiment 16)
Since the basic configuration of this embodiment is the same as that of the eleventh and thirteenth embodiments as shown in FIG. 19, the same reference numerals are given to common components, and the description thereof is omitted. Only the characteristic configuration will be described. In the present embodiment, both the acoustic-side feedback gain estimation means 11 and the line-side feedback gain estimation means 12 are provided, and the transmission signal voice section detection unit 37 and the reception signal voice section detection unit 38 each have a line-side feedback gain β. The transmitting side threshold value and the receiving side threshold value Ps0, δ used for determination of speech and non-speech are changed according to the estimated value | β ′ | and the estimated value | α ′ | of the acoustic side feedback gain α. There is a feature in the point. The configurations of the transmission signal voice section detection unit 37 and the reception signal voice section detection unit 38 are the same as those of the first embodiment shown in FIG.
[0109]
The estimated value | α ′ | of the acoustic feedback gain α is input to the received signal speech section detection unit 38, and if this estimated value | α ′ | is smaller than a predetermined reference value, the speech / non-speech determination unit The threshold value Ps0 to be compared with the instantaneous power estimate value Ps and the threshold value δ to be compared with the ratio Ps / Pn between the instantaneous power estimate value Ps and the background noise power estimate value Pn are set to predetermined optimum values. If the estimated value | α ′ | is equal to or greater than the reference value, the two threshold values Ps0 and δ are set to a value smaller than the optimum value. This makes it easier for the received signal voice section detection unit 38 to detect a voice section under the condition where the acoustic feedback gain α is relatively large, and the transmission bias mode is not set when a voice section is detected. Reception blocking is less likely to occur.
[0110]
On the other hand, an estimated value | β ′ | of the line side feedback gain β is inputted to the transmission signal speech section detecting unit 37, and if this estimated value | β ′ | is smaller than a predetermined reference value, speech / non- The threshold value Ps0 to be compared with the instantaneous power estimated value Ps and the threshold value δ to be compared with the ratio Ps / Pn between the instantaneous power estimated value Ps and the background noise power estimated value Pn are predetermined. If the estimated value | β ′ | is equal to or greater than the reference value, the two threshold values Ps0 and δ are set to values smaller than the optimum value. As a result, the voice signal section is easily detected by the transmission signal voice section detector 37 under a situation where the line-side feedback gain β is relatively large, and when the voice section is detected, the reception bias mode is not set. Transmission blocking is less likely to occur.
[0111]
In addition, instead of changing the two threshold values Ps0 and δ according to the estimated values | α ′ | and | β ′ | of the feedback gains α and β, the instantaneous power estimation that constitutes the speech section detection units 37 and 10 is performed. Even if the time constant used in the calculation in the unit and the background noise power estimation unit is changed, the same effect can be obtained. For example, if the estimated values | α ′ | and | β ′ | of the acoustic feedback gains α and β are smaller than a predetermined reference value, the time constant is set to a predetermined optimum value, and the estimated value | α ′ | , | Β ′ | is equal to or greater than the reference value, the time constant may be set to a value such that the rising characteristic is slower than the optimum value and the falling characteristic is faster.
[0112]
(Embodiment 17)
FIG. 20 shows a schematic configuration of a loudspeaker call terminal provided with the voice switching device VS of the present embodiment, and the same reference numerals are given to components common to the seventh embodiment.
[0113]
In the loudspeaker terminal according to the present embodiment, as in the sixth embodiment, the transmission side threshold and the transmission side time constant for the transmission signal voice section detection unit 37 and the reception side threshold for the reception signal voice section detection unit 38 are used. Value and transmission side time constant are the magnitude relationship between the estimated value | α '| of the acoustic side feedback gain α and the threshold value α0, and the magnitude relationship between the estimated value | β' | of the line side feedback gain β and the threshold value β0. Are stored in the flash memory 110 as a plurality of types (in this embodiment, two types) of values.
[0114]
In the loudspeaker terminal according to the present embodiment, as in the seventh embodiment, when starting a call, the CPU 200 transmits a call start request command to the DSP 100 via the communication interface and receives the call start request command. The command processing means 103 interprets the contents and uses the above parameters (transmission side threshold, transmission side time constant, reception side threshold, etc.) used in the transmission signal voice section detection unit 37 and the reception signal voice section detection unit 38. , Receiving side time constant) is given to the parameter setting means 104. Based on the received address data, the parameter setting unit 104 reads the threshold value and time constant data on the transmitting side and the receiving side from the flash memory 110. At this time, there are parameters whose values correspond to the magnitude relationship between the estimated value | α ′ | of the acoustic feedback gain α and the threshold value α0 and the magnitude relationship between the estimated value | β ′ | of the line feedback gain β and the threshold value β0. When selected, the parameter setting means 104 initializes the software module for realizing the voice section detection units 37 and 10. After these initial settings are completed, the parameter setting means 104 sets a flag indicating that the parameter initialization is completed, and the command processing means 103 transmits a response command to the CPU 200 in response to the setting of the flag.
[0115]
Thus, in the present embodiment, similarly to the seventh embodiment, the difference between the transmission characteristics on the near end side and the transmission characteristics on the far end side can be easily corrected, and transmission blocking and reception blocking can be made difficult to occur.
[0116]
(Embodiment 18)
FIG. 21 shows a schematic configuration of a loudspeaker call terminal provided with the voice switching device VS of the present embodiment, and the same components as those in the eighth embodiment are denoted by the same reference numerals.
[0117]
In the present embodiment, the transmission side threshold value and the transmission side time constant for the transmission signal voice section detection unit 37, and the reception side threshold value and the transmission side time constant for the reception signal voice section detection unit 38 are: A plurality of types according to the magnitude relationship between the estimated value | α ′ | of the acoustic side feedback gain α and the threshold value α0 and the magnitude relationship between the estimated value | β ′ | of the line side feedback gain β and the threshold value β0 (in this embodiment) In addition to being set to two types of values, different values can be set according to the type of the other party's call terminal (for example, door phone and lobby intercom). Specifically, as in the eighth embodiment, the doorphone parameters (threshold values and time constants on the transmitting side and the receiving side) and the lobby interphone parameters are stored in the two areas M1 and M2 of the flash memory 110, respectively. is doing. Then, the CPU 200 designates the address of the parameter according to the type of the call terminal of the call partner with a command, and based on the address data received by the parameter setting means 104, the transmitter and the call receiver according to the type of the call terminal of the other party Side threshold value and time constant data are read from the areas M1 and M2 of the flash memory 110. At this time, similarly to the seventeenth embodiment, the magnitude relationship between the estimated value | α ′ | of the acoustic side feedback gain α and the threshold value α0, and the magnitude relationship between the estimated value | β ′ | of the line side feedback gain β and the threshold value β0. A parameter having a corresponding value is selected, and is initially set by the parameter setting means 104 for the software module that implements the speech section detection units 37 and 10.
[0118]
Thus, in the present embodiment, the threshold values and time constants on the transmitting side and the receiving side for the transmission signal voice section detection unit 37 and the reception signal voice section detection unit 38 are external (CPU200) for each call terminal on the other side. Therefore, it is possible to easily correct a difference between two transfer characteristics that differ for each call terminal on the other side, and to prevent transmission blocking and reception blocking.
[0119]
(Embodiment 19)
In this embodiment, the bias mode control means 8 is configured so that the difference between the near-end side noise power estimated value PNn and the far-end side noise power estimated value PFn and the feedback gain estimated value | α ′ | , | Β ′ |, the gain GT and GR of the transmission bias mode setting amplifier 6 and the reception bias mode setting amplifier 7 are adaptively updated. However, since the entire configuration is the same as that of the fifteenth embodiment, illustration and description thereof are omitted.
[0120]
When the transmission mode is set to the transmission bias mode, the bias mode control unit 8 increases the difference between the near-end side noise power estimation value PNn and the far-end side noise power estimation value PFn (= PFn−PNn). When the gain GT of the mode setting amplifier 6 is increased to set the reception bias mode, the difference (= PNn−PFn) between the estimated value PNn of the near-end side noise power and the estimated value PFn of the far-end side noise power is large. Control is performed so that the gain GR of the reception bias mode setting amplifier 7 is reduced. Here, when the gain GT of the transmission bias mode setting amplifier 6 is increased, on the condition that the estimated value | α ′ | of the acoustic side feedback gain α is lower than the threshold value α0, it is used for setting the reception bias mode. When the gain GR of the amplifier 7 is increased, it is a condition that the estimated value | β ′ | of the line side feedback gain β is below the threshold value β0. Further, both the threshold values α0 and β0 are increased or decreased so that the product of the gain GT and the threshold value α0 and the product of the gain GR and the threshold value β0 are both substantially constant as the gains GT and GR are increased or decreased. In addition, when updating the values of both gains GT and GR, it is a condition that each of the voice section detection units 37 and 38 of the transmission signal and the reception signal has not detected the voice section.
[0121]
As described above, the bias mode control means 8 determines the difference between the near-end side noise power estimated value PNn and the far-end side noise power estimated value PFn, and the acoustic-side and line-side feedback gain estimated values | α ′ |, | Β ′ | is used to adaptively update the gains GT and GR of the transmission bias mode setting amplifier 6 and the reception bias mode setting amplifier 7. It is possible to switch the call mode to the transmission mode or the reception mode in a balanced manner without causing any interruption, and to further reduce the far end side and near end side utterance levels necessary for switching the call mode.
[0122]
By the way, it is desirable to set upper limit values for the gains GT and GR of the transmission bias mode setting amplifier 6 and the reception bias mode setting amplifier 7 which are adaptively updated by the bias mode control means 8 as described above. That is, if both gains GT and GR are set to a value larger than necessary, reception blocking and transmission blocking will occur. Therefore, an upper limit value is set in advance for gains GT and GR, and the bias mode control is performed. By preventing the upper limit from being exceeded when the means 8 adaptively updates both gains GT and GR, occurrence of reception blocking and transmission blocking can be suppressed.
[0123]
【The invention's effect】
The invention according to claim 1 is used for the above-mentioned loudspeaker call terminal of the loudspeaker call system in which a loudspeaker call terminal having a microphone and a speaker is connected to another call terminal or the loudspeaker call terminal by wire, and to transmit sound collected by the microphone Transmission side loss insertion means for inserting loss into the transmission side signal path for transmitting the signal to the line, and reception for inserting loss into the reception side signal path for transmitting the reception signal received from the line to the speaker Side loss insertion means, transmission deviation mode setting amplification means for taking out and amplifying the transmission signal input to the transmission side loss insertion means, and reception signal input to the reception side loss insertion means Amplifying means for setting the reception bias mode to amplify, a transmission mode and a receiving signal amplified by the transmission bias mode setting amplifying means and the reception bias mode setting amplifying means. Insertion loss for switching the call mode between the transmission mode and the reception mode by controlling the amount of loss inserted into the path by the transmission side loss insertion means and the reception side loss insertion means according to the estimation result Amount control means, near-end side noise power estimation means for estimating near-end side noise power included in the transmitted signal, and far-end side noise power estimation means for estimating far-end side noise power included in the received signal And a bias mode control means for adjusting the gains of the transmission bias mode setting amplification means and the reception bias mode setting amplification means in accordance with the estimated values of the far-end side noise power and the near-end side noise power. The insertion loss amount control means estimates the instantaneous power of the input signal to the transmission side loss insertion means, and the instantaneous power of the input signal to the reception side loss insertion means. A second instantaneous power estimation unit to be determined, and a feedback from the input point to the transmission side loss insertion means to the input point to the reception side loss insertion means through the transmission side loss insertion means and the wraparound on the line side A line feedback gain multiplier having a value determined according to the gain of the system to be used as a coefficient, and the transmission side from the input point to the reception side loss insertion means through the reception side loss insertion means and the wraparound on the acoustic side The output signal of the acoustic feedback gain multiplier having a coefficient determined by the gain of the path to the input point to the side loss insertion means and the output signal of the second instantaneous power estimator are input to the acoustic feedback gain multiplier. The first comparator for comparing the magnitude relationship between the output signal obtained in this way and the output signal of the first instantaneous power estimator, and the output signal of the first instantaneous power estimator are input to the line feedback gain multiplier. Output signal and second instantaneous power estimator A second comparator for comparing the magnitude relationship with the output signal of the received signal, a transmission signal voice section detecting unit for detecting a voice section of the transmitted signal, a received signal voice section detecting unit for detecting the voice section of the received signal, The communication state is determined based on the comparison results of the first comparator and the second comparator and the detection results of the transmission signal voice section detection unit and the reception signal voice section detection unit, and the transmission side loss insertion means and the reception are received. An insertion loss amount distribution processing unit for controlling an insertion loss amount of the side loss insertion means, and the bias mode control means is configured to determine a relationship between magnitudes of estimated values of noise power on the far end side and the near end side, and the transmission Based on the detection result of the signal voice section detector, the gain of the neutral mode and the amplification unit for setting the reception bias mode that makes the gains of the transmission bias mode setting amplification means and the reception bias mode setting amplification means substantially equal to each other The transmission bias mode The reception mode is set to any one of the receiving decentering modes that increase more than the gain of the amplifying means, and the decentering mode is set based on only the estimated values of the near-end side noise power and the far-end side noise power as in the past. Instead, the bias mode control means sets the bias mode based on the magnitude relationship between the estimated values of the noise power on the far end side and the near end side and the detection result of the transmission signal voice section detector. When the near-end side is speaking unilaterally, the biased mode control means does not set the received weighted mode. Can be prevented from being accidentally switched to the reception mode, and voice interruptions and unnatural inflections can be suppressed.
[0124]
According to a second aspect of the present invention, in the first aspect of the invention, the bias mode control means is configured such that the estimated value of the near-end side noise power is equal to the estimated value of the far-end side noise power and a predetermined first noise power ratio coefficient. When the duration of the state in which no speech section is detected by the transmission signal speech section detecting unit is equal to or longer than the first predetermined time, the gain of the reception bias mode setting amplifier is transmitted. It is characterized in that it is set to the reception deviation mode by increasing the gain of the amplification means for setting the deviation mode, and the characteristic of switching to the reception deviation mode by the deviation mode control means according to the first noise power ratio coefficient and the first predetermined time. Can be set arbitrarily.
[0125]
In order to achieve the above object, the invention according to claim 3 is used for the above-mentioned loudspeaker call terminal of a loudspeaker call system in which a loudspeaker call terminal having a microphone and a speaker is connected to another call terminal or a loudspeaker call terminal by wire, Transmission side loss insertion means for inserting a loss into a transmission side signal path for transmitting a transmission signal collected by the microphone to the line, and a reception side for transmitting a reception signal received from the line to the speaker A receiver-side loss insertion unit that inserts a loss into a signal path; a transmission bias mode setting amplification unit that extracts and amplifies a transmission signal input to the transmission-side loss insertion unit; and the reception-side loss insertion unit. A receiving bias mode setting amplifying means for picking up and amplifying the received receiving signal; a transmission bias mode setting amplifying means; and a transmission signal amplified by the receiving bias mode setting amplifying means. Estimating the call mode based on the received signal and controlling the amount of loss inserted into the path by the transmission side loss insertion means and the reception side loss insertion means in accordance with the estimation result, thereby changing the call mode to the transmission mode. Insertion loss control means for switching to the reception mode, near-end noise power estimation means for estimating the near-end noise power included in the transmitted signal, and far-end noise power for estimating the far-end noise power included in the received signal End-side noise power estimation means, and a bias that adjusts the gains of the transmission bias mode setting amplification means and the reception bias mode setting amplification means according to the estimated values of the far-end side noise power and the near-end side noise power Mode control means, wherein the insertion loss amount control means estimates the instantaneous power of the input signal to the transmission side loss insertion means, and the reception side loss insertion means A second instantaneous power estimator for estimating the instantaneous power of the input signal of the receiver, and the insertion of the receiving side loss from the input point to the transmitting side loss insertion means through the transmission side loss insertion means and the wraparound on the line side A line feedback gain multiplier having as a coefficient a value determined according to the gain of the system that feeds back to the input point to the input unit, and from the input point to the reception side loss insertion unit to the reception side loss insertion unit and the acoustic side And an acoustic feedback gain multiplier having a value determined according to the gain of the path reaching the input point to the transmission side loss insertion means via the wraparound and the output signal of the second instantaneous power estimator The first comparator for comparing the magnitude relationship between the output signal obtained by input to the acoustic feedback gain multiplication unit and the output signal of the first instantaneous power estimation unit, and the output signal of the first instantaneous power estimation unit Output signal obtained by input to feedback gain multiplier A second comparator for comparing the magnitude relationship between the signal and the output signal of the second instantaneous power estimation unit, a transmission signal speech segment detection unit for detecting a speech segment of the transmitted signal, and a speech segment of the received signal. The call state is determined based on the detected reception signal voice interval detection unit, the comparison results of the first comparator and the second comparator, and the detection results of the transmission signal voice interval detection unit and the reception signal voice interval detection unit. An insertion loss amount distribution processing unit for controlling the insertion loss amount of the transmission side loss insertion means and the reception side loss insertion means, and the bias mode control means is configured to control each noise power of the far end side and the near end side. Based on the magnitude relationship of the estimated values and the detection result of the reception signal voice section detection unit, the neutral mode for substantially equalizing the gains of the transmission bias mode setting amplification unit and the reception bias mode setting amplification unit and the transmission mode Of the amplification means for setting the talk bias mode Each of the estimated values of the near-end side noise power and the far-end side noise power as in the prior art. Instead of setting the bias mode based only on the load mode, the bias mode control means is based on the magnitude relationship of the estimated values of the noise power on the far end side and the near end side and the detection result of the received signal speech section detection unit. For example, when the unilateral voice is uttered from the far end, the bias mode control means does not set the transmission bias mode, so it is inserted even if the far end speaker is uttering. It is possible to prevent the loss amount control means from erroneously switching from the reception mode to the transmission mode, and it is possible to suppress voice interruption and unnatural inflection.
[0126]
According to a fourth aspect of the present invention, in the third aspect of the present invention, the bias mode control means is configured such that the estimated value of the far-end side noise power is equal to the estimated value of the near-end side noise power and a predetermined second noise power ratio coefficient. And the gain of the transmission bias mode setting amplification means when the duration of the state in which no voice section is detected by the reception signal voice section detector becomes a second predetermined time or more. Switching to the transmission bias mode by the bias mode control means according to the second noise power ratio coefficient and the second predetermined time is set to the transmission bias mode by increasing the gain of the mode setting amplification means. Characteristics can be set arbitrarily.
[0127]
According to a fifth aspect of the present invention, in the invention according to any one of the first to fourth aspects, the transmission side or reception side voice section detection unit is included in an input signal to the transmission side or reception side loss insertion means. A noise power estimator for estimating noise power; and comparing the first or second instantaneous power estimated value with a predetermined threshold value and a product of the noise power estimated value and the predetermined value. A speech / non-speech determination unit that determines that the input signal to the loss insertion means is a speech signal when it is greater than the threshold and the instantaneous power estimation value is greater than the product, and determines that it is a non-speech signal when it is not greater The threshold value and the predetermined value can be set from the outside, and can be easily adapted to various voice call terminals.
[0128]
The invention of claim 6 is the invention according to any one of claims 1 to 4, wherein the voice section detecting unit on the transmitting side or the receiving side is included in an input signal to the loss insertion means on the transmitting side or the receiving side. A noise power estimator for estimating noise power; and comparing the first or second instantaneous power estimated value with a predetermined threshold value and a product of the noise power estimated value and the predetermined value. A speech / non-speech determination unit that determines that the input signal to the loss insertion means is a speech signal when it is greater than the threshold and the instantaneous power estimation value is greater than the product, and determines that it is a non-speech signal when it is not greater And the instantaneous power estimator comprises a filter having a steep rise and a gradual fall, and the noise power estimator has a slow rise and a steep fall. It consists filter having, characterized in that the settable parameters that determine the characteristics of the filter from the outside, can be readily adapted to a variety of hands-free communication device.
[0129]
The invention of claim 7 is the invention of claim 5 or 6, wherein the threshold value and the predetermined value or the parameter can be individually set from the outside to the voice section detecting unit on the transmitting side or the receiving side. The difference between the near-end-side transfer characteristic and the far-end-side transfer characteristic can be easily corrected.
[0130]
The invention of claim 8 is characterized in that, in the invention of claim 5 or 6, the threshold value and the predetermined value or the parameter can be set to different values depending on the type of the other call terminal, The difference between the near-end-side transfer characteristic and the far-end-side transfer characteristic that is different for each call terminal on the other side can be easily corrected.
[0131]
The invention according to claim 9 is the invention according to any one of claims 1 to 8, wherein the voice section detecting unit on the transmitting side or the receiving side is included in an input signal to the loss insertion means on the transmitting side or the receiving side. A noise power estimator for estimating noise power; and comparing the first or second instantaneous power estimated value with a predetermined threshold value and a product of the noise power estimated value and the predetermined value. A speech / non-speech determination unit that determines that the input signal to the loss insertion means is a speech signal when it is greater than the threshold and the instantaneous power estimation value is greater than the product, and determines that it is a non-speech signal when it is not greater The near-end side noise power estimation means is also used as the noise power estimation section constituting the transmission-side voice section detection section, and the far-end side noise power estimation means is used as the reception-side voice section detection. Parts and characterized in that it is combined with the noise power estimating section constituting the configuration can be simplified.
[0132]
According to a tenth aspect of the present invention, in the second aspect of the present invention, the deviation mode control means is configured such that the estimated value of the near-end side noise power is the far end even if the duration is less than the first predetermined time. A predetermined third predetermined condition in which a state that is equal to or greater than a product of the estimated value of the side noise power and a predetermined third noise power ratio coefficient larger than the first noise power ratio coefficient is longer than the first predetermined time. It is characterized in that it is set to the reception bias mode when it continues for more than a time, and it is difficult to set the reception bias mode even if there is a high level of unsteady noise around the near-end call terminal. Can be prevented from occurring in a so-called one-sided fall that is fixed in the transmission mode.
[0133]
According to an eleventh aspect of the present invention, in the invention according to the fourth aspect, the deviation mode control means has the estimated value of the far end side noise power as the near end even if the duration is less than the second predetermined time. A predetermined fourth predetermined condition in which a state in which the estimated value of the side noise power and a predetermined fourth noise power ratio coefficient larger than the second noise power ratio coefficient is equal to or greater than the second predetermined time is longer It is characterized in that it is set to the transmission eccentric mode when it continues for more than a time, and it is difficult to set the transmission eccentric mode even if there is a high level of unsteady noise around the far-end telephone terminal. It is possible to prevent the so-called half-falling that the call mode is fixed to the reception mode.
[0134]
The invention of claim 12 is claimed in claim 3 The acoustic side feedback gain for estimating the acoustic side feedback gain of the path returning from the output point of the reception side loss insertion means to the input point of the transmission side loss insertion means via the acoustic echo path on the near end side The deviation mode control means includes an estimation means, and the estimated value of the acoustic feedback gain does not shift to the transmission deviation mode unless the estimated value of the acoustic side feedback gain satisfies a predetermined condition, and can prevent reception blocking.
[0135]
According to a thirteenth aspect of the present invention, in the twelfth aspect of the invention, the bias mode control means is configured such that the estimated value of the far-end side noise power is an estimated value of the near-end side noise power and the second noise power ratio coefficient. Or when the detection result of the received signal speech section detection unit is a non-speech section and the estimated value of the acoustic feedback gain is less than a predetermined threshold value continues for a predetermined time or more. When the estimated value of the acoustic feedback gain is equal to or greater than a predetermined threshold, the eccentric mode control means always sets the neutral mode regardless of other conditions, and the acoustic feedback is set. Since the transmission bias mode is not set in a situation where the gain is relatively large, reception blocking can be prevented. In addition, since the threshold value can be used as a variable parameter, the performance of the device can be set variously according to the installation environment of the device using the voice switching device.
[0136]
The invention of claim 14 is claimed in claim 1 The line-side feedback gain for estimating the line-side feedback gain of the path that returns from the output point of the transmitting-side loss insertion means to the input point of the receiving-side loss insertion means via the far-end line echo path The deviation mode control means is provided with an estimation means, and is characterized by not shifting to the reception deviation mode unless the estimated value of the line-side feedback gain satisfies a predetermined condition, and can prevent transmission blocking.
[0137]
The invention according to a fifteenth aspect is the invention according to the fourteenth aspect, wherein the bias mode control means is configured such that the estimated value of the near-end side noise power is an estimated value of the far-end side noise power and the first noise power ratio coefficient. When the detection result of the transmission signal voice section detection unit is a non-voice section and the estimated value of the line-side feedback gain is less than a predetermined threshold value continues for a predetermined time or more If the estimated value of the line-side feedback gain is equal to or greater than a predetermined threshold, the bias mode control means always sets the neutral mode regardless of other conditions, and the line-side feedback mode is set. In a situation where the gain is relatively large, since the reception bias mode is not set, transmission blocking can be prevented. In addition, since the threshold value can be used as a variable parameter, the performance of the device can be set variously according to the installation environment of the device using the voice switching device.
[0138]
According to a sixteenth aspect of the present invention, in the twelfth aspect of the present invention, a line for a path that returns from the output point of the transmission side loss insertion means to the input point of the reception side loss insertion means via a line echo path on the far end side. Line-side feedback gain estimation means for estimating the side feedback gain, and the bias mode control means does not shift to the reception bias mode if the estimated value of the line-side feedback gain does not satisfy a predetermined condition. Each speech section detector on the receiver side includes a noise power estimator for estimating a noise power included in an input signal to the loss insertion means on the transmitter side or the receiver side, and the first or second instantaneous power estimate value. Loss insertion means for comparing a predetermined threshold and a product of the noise power estimated value and the predetermined value, respectively, and when the instantaneous power estimated value is larger than the threshold and the instantaneous power estimated value is larger than the product A speech / non-speech determination unit that determines that the input signal is a speech signal and determines that the input signal is a non-speech signal when the input signal is not large, and each speech section detection unit on the transmitting side and the receiving side includes the acoustic side and The threshold value and the predetermined value are changed in accordance with the estimated value of each feedback gain estimating means on the line side, and the voice section is detected by the voice section detecting unit under a situation where the feedback gain is relatively large. It becomes easy, and when the voice section is detected, since the bias mode is not set, transmission or reception blocking is less likely to occur.
[0139]
According to a seventeenth aspect of the present invention, in the twelfth aspect of the present invention, a line on a path for returning from the output point of the transmission side loss insertion means to the input point of the reception side loss insertion means via a line echo path on the far end side. Line-side feedback gain estimation means for estimating the side feedback gain, and the bias mode control means sets the reception bias mode when the estimated value of the line-side feedback gain satisfies a predetermined condition, Each speech section detector on the receiver side includes a noise power estimator for estimating a noise power included in an input signal to the loss insertion means on the transmitter side or the receiver side, and the first or second instantaneous power estimate value. Loss insertion means for comparing a predetermined threshold and a product of the noise power estimated value and the predetermined value, respectively, and when the instantaneous power estimated value is larger than the threshold and the instantaneous power estimated value is larger than the product What A voice / non-speech determination unit that determines an input signal as a speech signal and a non-speech signal when the input signal is not large, and the instantaneous power estimation unit is a filter having a characteristic that the rise is steep and the fall is gradual. The noise power estimation unit is composed of a filter having a characteristic that the rising edge is gradual and the falling edge is steep, and the speech section detection units on the transmitting side and the receiving side estimate the feedback gains on the acoustic side and the line side. The parameter for determining the characteristic is changed according to the estimated value of the means, and the voice section is easily detected by the voice section detection unit under a situation where the feedback gain is relatively large, and the voice section is detected. In this case, since the bias mode is not set, transmission or reception blocking is less likely to occur.
[0140]
In the invention of claim 18, in the invention of claim 17, the threshold value, the predetermined value, and the parameter can be individually set from the outside for the respective voice section detection units on the transmitting side and the receiving side. The difference between the transmission characteristic on the near end side and the transmission characteristic on the far end side can be easily corrected, and transmission or reception blocking is less likely to occur.
[0141]
The invention of claim 19 is characterized in that, in the invention of claim 17, the threshold value, the predetermined value and the parameter can be set to different values according to the type of the other call terminal. It is possible to easily correct the difference between the two transfer characteristics of the near-end side and the far-end side, which are different for each of the call terminals, and to prevent transmission blocking and reception blocking.
[0142]
According to a twentieth aspect of the present invention, in the twelfth aspect of the present invention, a line on a path for returning from the output point of the transmission side loss insertion means to the input point of the reception side loss insertion means via a line echo path on the far end side. Line-side feedback gain estimating means for estimating a side-side feedback gain, and the deviation mode control means does not shift to a reception-side deviation mode if the estimated value of the line-side feedback gain does not satisfy a predetermined condition. Referring to the difference between the estimated power value and the estimated far-end noise power, and the feedback gain estimated values on the acoustic side and the line side, the transmission bias mode setting amplification means and the reception bias mode setting amplification It is characterized in that the gain of the means is adaptively updated, and it is possible to switch the call mode to the transmission mode or the reception mode in a balanced manner without causing unnatural interruption of words during the call. It is possible to lower the utterance level of the far-end and near-end-side necessary for switching the mode.
[0143]
According to a twenty-first aspect of the present invention, in the twentieth aspect of the invention, an upper limit value is set for the gain of the transmission bias mode setting amplification unit and the reception bias mode setting amplification unit that are adaptively updated by the bias mode control unit. It is possible to suppress the occurrence of reception blocking and transmission blocking by preventing the deviation mode control means from updating the upper limit value when both gains are adaptively updated.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a first embodiment.
FIG. 2 is a block diagram of transmission signal voice section detecting means in the same as above.
FIG. 3 is a flowchart for explaining the operation of the deflection mode control means in the same as above.
FIG. 4 is a block diagram of a bias mode control unit according to the second embodiment.
FIG. 5 is a block diagram showing a third embodiment.
FIG. 6 is a flowchart for explaining the operation of the deflection mode control means in the above.
FIG. 7 is a block diagram of a bias mode control unit according to the fourth embodiment.
FIG. 8 is a partially omitted block diagram illustrating a loudspeaker call terminal including a voice switching device according to a fifth embodiment.
FIG. 9 is a partially omitted block diagram illustrating a loudspeaker call terminal including a voice switching device according to a sixth embodiment.
FIG. 10 is a partially omitted block diagram illustrating a loudspeaker call terminal including a voice switching device according to a seventh embodiment.
FIG. 11 is a partially omitted block diagram illustrating a loudspeaker call terminal including a voice switching device according to an eighth embodiment.
12 is a block diagram showing Embodiment 9. FIG.
13 is a block diagram of a bias mode control means in Embodiment 10. FIG.
14 is a block diagram of a bias mode control means in Embodiment 11. FIG.
15 is a block diagram showing Embodiment 12. FIG.
FIG. 16 is a block diagram of a bias mode control means in the thirteenth embodiment.
FIG. 17 is a block diagram showing a fourteenth embodiment.
FIG. 18 is a block diagram of the bias mode control means in the fifteenth embodiment.
19 is a block diagram showing Embodiment 16. FIG.
FIG. 20 is a partially omitted block diagram illustrating a loudspeaker call terminal including a voice switching device according to a seventeenth embodiment.
FIG. 21 is a partially omitted block diagram illustrating a loudspeaker call terminal including a voice switching device according to an eighteenth embodiment.
FIG. 22 is a block diagram showing a conventional example.
[Explanation of symbols]
1 Transmitting side loss insertion means
2 Loss insertion means on the receiver side
3 Insertion loss control means
4 Near-end noise power estimation means
5 Far-end noise power estimation means
6 Transmitter bias mode setting amplifier
7 Amplifier for setting the reception bias mode
8 Unbalance mode control means
37 Transmission signal voice section detector

Claims (21)

マイクロホン及びスピーカを有する拡声通話端末が他の通話端末又は拡声通話端末に有線で接続される拡声通話系の前記拡声通話端末に用いられ、前記マイクロホンで集音する送話信号を回線へ伝送するための送話側信号経路に損失を挿入する送話側損失挿入手段と、回線から受信した受話信号を前記スピーカへ伝送するための受話側信号経路に損失を挿入する受話側損失挿入手段と、前記送話側損失挿入手段に入力される送話信号を取り出して増幅する送話偏重モード設定用増幅手段と、前記受話側損失挿入手段に入力される受話信号を取り出して増幅する受話偏重モード設定用増幅手段と、前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段で増幅された送話信号及び受話信号に基づいて通話モードを推定するとともに該推定結果に応じて前記送話側損失挿入手段並びに受話側損失挿入手段が前記経路に挿入する損失量を制御して前記通話モードを送話モードと受話モードに切り換える挿入損失量制御手段と、送話信号に含まれる近端側の雑音パワーを推定する近端側雑音パワー推定手段と、受話信号に含まれる遠端側の雑音パワーを推定する遠端側雑音パワー推定手段と、遠端側雑音パワー並びに近端側雑音パワーの各推定値に応じて前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の各利得を調整する偏重モード制御手段とを備え、前記挿入損失量制御手段が、前記送話側損失挿入手段への入力信号の瞬時パワーを推定する第１の瞬時パワー推定部と、前記受話側損失挿入手段への入力信号の瞬時パワーを推定する第２の瞬時パワー推定部と、前記送話側損失挿入手段への入力点から前記送話側損失挿入手段並びに回線側での回り込みを経て前記受話側損失挿入手段への入力点へ帰還する系の利得に応じて決定される値を係数にもつ回線帰還利得乗算部と、前記受話側損失挿入手段への入力点から前記受話側損失挿入手段並びに音響側での回り込みを経て前記送話側損失挿入手段への入力点へ到る経路の利得に応じて決定される値を係数にもつ音響帰還利得乗算部と、第２の瞬時パワー推定部の出力信号を音響帰還利得乗算部へ入力して得られる出力信号と第１の瞬時パワー推定部の出力信号との大小関係を比較する第１の比較器と、第１の瞬時パワー推定部の出力信号を回線帰還利得乗算部へ入力して得られる出力信号と第２の瞬時パワー推定部の出力信号との大小関係を比較する第２の比較器と、送話信号の音声区間を検出する送話信号音声区間検出部と、受話信号の音声区間を検出する受話信号音声区間検出部と、第１の比較器及び第２の比較器の比較結果と送話信号音声区間検出部及び受話信号音声区間検出部の検出結果とに基づいて通話状態を判定し送話側損失挿入手段及び受話側損失挿入手段の挿入損失量を制御する挿入損失量分配処理部とを具備し、前記偏重モード制御手段は、遠端側並びに近端側の各雑音パワーの推定値の大小関係と前記送話信号音声区間検出部の検出結果に基づいて、前記送話偏重モード設定用増幅手段及び前記受話偏重モード設定用増幅手段の各利得をほぼ等しくする中立モードと前記受話偏重モード設定用増幅手段の利得を前記送話偏重モード設定用増幅手段の利得よりも増大させる受話偏重モードの何れかに設定することを特徴とする音声切換装置。  A voice call terminal having a microphone and a speaker is used for the voice call terminal of the voice call system in which the voice call terminal is connected to another call terminal or the voice call terminal by wire to transmit a transmission signal collected by the microphone to the line. A transmission side loss insertion means for inserting loss into the transmission side signal path, a reception side loss insertion means for inserting loss into the reception side signal path for transmitting the reception signal received from the line to the speaker, and A transmission bias mode setting amplification unit that extracts and amplifies the transmission signal input to the transmission side loss insertion unit, and a reception bias mode setting unit that extracts and amplifies the reception signal input to the reception side loss insertion unit. The communication mode is estimated based on the amplification means, the transmission signal and the reception signal amplified by the transmission deviation mode setting amplification means and the reception deviation mode setting amplification means. And an insertion loss amount control means for controlling the amount of loss inserted into the path by the transmission side loss insertion means and the reception side loss insertion means in accordance with the estimation result to switch the call mode between the transmission mode and the reception mode. A near-end side noise power estimating means for estimating the near-end side noise power included in the transmitted signal; a far-end side noise power estimating means for estimating the far-end side noise power included in the received signal; A bias mode control means for adjusting the gains of the transmission bias mode setting amplification means and the reception bias mode setting amplification means in accordance with the estimated values of the side noise power and the near-end side noise power, and the insertion loss. A quantity control means for estimating the instantaneous power of the input signal to the transmission side loss insertion means; and a second instantaneous power estimation section for estimating the instantaneous power of the input signal to the reception side loss insertion means. Instant Depending on the gain of the system that returns from the input point to the transmission side loss insertion means to the input point to the reception side loss insertion means through the transmission side loss insertion means and the wraparound on the line side. A line feedback gain multiplication unit having a value determined in accordance with a coefficient, and an input point to the reception-side loss insertion means to the reception-side loss insertion means and a wraparound on the acoustic side to the transmission-side loss insertion means An output signal obtained by inputting the output signal of the acoustic feedback gain multiplier having a value determined according to the gain of the path to the input point and the second instantaneous power estimator to the acoustic feedback gain multiplier And a first comparator for comparing the magnitude relationship between the output signal of the first instantaneous power estimation unit and the output signal obtained by inputting the output signal of the first instantaneous power estimation unit to the line feedback gain multiplier The magnitude of the output signal of the second instantaneous power estimator A second comparator for comparing the relationship, a transmission signal voice section detector for detecting a voice section of the transmission signal, a received signal voice section detector for detecting a voice section of the received signal, and a first comparator And determining the call state based on the comparison results of the second comparator and the detection results of the transmission signal voice section detection unit and the reception signal voice section detection unit, and inserting the transmission side loss insertion means and the reception side loss insertion means. An insertion loss amount distribution processing unit for controlling a loss amount, and the bias mode control means includes a magnitude relationship between estimated values of noise power on the far end side and the near end side and a transmission signal speech section detection unit. Based on the detection result, the gain of the neutralization mode for substantially equalizing each gain of the transmission bias mode setting amplification means and the reception bias mode setting amplification means and the gain of the reception bias mode setting amplification means are set to the transmission bias mode. Gain of setting amplification means Voice switching apparatus characterized by setting either of the receiving unbalance mode in which remote increases. 前記偏重モード制御手段は、前記近端側雑音パワーの推定値が前記遠端側雑音パワーの推定値と所定の第１の雑音パワー比係数との積以上の値となり且つ前記送話信号音声区間検出部によって音声区間が検出されない状態の継続時間が第１の所定時間以上となったときに受話偏重モード設定用増幅手段の利得を送話偏重モード設定用増幅手段の利得よりも増大させて受話偏重モードに設定することを特徴とする請求項１記載の音声切換装置。  The bias mode control means is configured such that the estimated value of the near-end side noise power is equal to or greater than a product of the estimated value of the far-end side noise power and a predetermined first noise power ratio coefficient, and the transmission signal speech section When the duration of the state in which the voice section is not detected by the detector becomes equal to or longer than the first predetermined time, the gain of the reception bias mode setting amplification means is increased more than the gain of the transmission bias mode setting amplification means. 2. The voice switching device according to claim 1, wherein the voice switching device is set to a bias mode. マイクロホン及びスピーカを有する拡声通話端末が他の通話端末又は拡声通話端末に有線で接続される拡声通話系の前記拡声通話端末に用いられ、前記マイクロホンで集音する送話信号を回線へ伝送するための送話側信号経路に損失を挿入する送話側損失挿入手段と、回線から受信した受話信号を前記スピーカへ伝送するための受話側信号経路に損失を挿入する受話側損失挿入手段と、前記送話側損失挿入手段に入力される送話信号を取り出して増幅する送話偏重モード設定用増幅手段と、前記受話側損失挿入手段に入力される受話信号を取り出して増幅する受話偏重モード設定用増幅手段と、前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段で増幅された送話信号及び受話信号に基づいて通話モードを推定するとともに該推定結果に応じて前記送話側損失挿入手段並びに受話側損失挿入手段が前記経路に挿入する損失量を制御して前記通話モードを送話モードと受話モードに切り換える挿入損失量制御手段と、送話信号に含まれる近端側の雑音パワーを推定する近端側雑音パワー推定手段と、受話信号に含まれる遠端側の雑音パワーを推定する遠端側雑音パワー推定手段と、遠端側雑音パワー並びに近端側雑音パワーの各推定値に応じて前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の各利得を調整する偏重モード制御手段とを備え、前記挿入損失量制御手段が、前記送話側損失挿入手段への入力信号の瞬時パワーを推定する第１の瞬時パワー推定部と、前記受話側損失挿入手段への入力信号の瞬時パワーを推定する第２の瞬時パワー推定部と、前記送話側損失挿入手段への入力点から前記送話側損失挿入手段並びに回線側での回り込みを経て前記受話側損失挿入手段への入力点へ帰還する系の利得に応じて決定される値を係数にもつ回線帰還利得乗算部と、前記受話側損失挿入手段への入力点から前記受話側損失挿入手段並びに音響側での回り込みを経て前記送話側損失挿入手段への入力点へ到る経路の利得に応じて決定される値を係数にもつ音響帰還利得乗算部と、第２の瞬時パワー推定部の出力信号を音響帰還利得乗算部へ入力して得られる出力信号と第１の瞬時パワー推定部の出力信号との大小関係を比較する第１の比較器と、第１の瞬時パワー推定部の出力信号を回線帰還利得乗算部へ入力して得られる出力信号と第２の瞬時パワー推定部の出力信号との大小関係を比較する第２の比較器と、送話信号の音声区間を検出する送話信号音声区間検出部と、受話信号の音声区間を検出する受話信号音声区間検出部と、第１の比較器及び第２の比較器の比較結果と送話信号音声区間検出部及び受話信号音声区間検出部の検出結果とに基づいて通話状態を判定し送話側損失挿入手段及び受話側損失挿入手段の挿入損失量を制御する挿入損失量分配処理部とを具備し、前記偏重モード制御手段は、遠端側並びに近端側の各雑音パワーの推定値の大小関係と前記受話信号音声区間検出部の検出結果に基づいて、前記送話偏重モード設定用増幅手段及び前記受話偏重モード設定用増幅手段の各利得をほぼ等しくする中立モードと前記送話偏重モード設定用増幅手段の利得を前記受話偏重モード設定用増幅手段の利得よりも増大させる送話偏重モードの何れかに設定することを特徴とする音声切換装置。  A voice call terminal having a microphone and a speaker is used for the voice call terminal of the voice call system in which the voice call terminal is connected to another call terminal or the voice call terminal by wire to transmit a transmission signal collected by the microphone to the line. A transmission side loss insertion means for inserting loss into the transmission side signal path, a reception side loss insertion means for inserting loss into the reception side signal path for transmitting the reception signal received from the line to the speaker, and A transmission bias mode setting amplification unit that extracts and amplifies the transmission signal input to the transmission side loss insertion unit, and a reception bias mode setting unit that extracts and amplifies the reception signal input to the reception side loss insertion unit. The communication mode is estimated based on the amplification means, the transmission signal and the reception signal amplified by the transmission deviation mode setting amplification means and the reception deviation mode setting amplification means. And an insertion loss amount control means for controlling the amount of loss inserted into the path by the transmission side loss insertion means and the reception side loss insertion means in accordance with the estimation result to switch the call mode between the transmission mode and the reception mode. A near-end side noise power estimating means for estimating the near-end side noise power included in the transmitted signal; a far-end side noise power estimating means for estimating the far-end side noise power included in the received signal; A bias mode control means for adjusting the gains of the transmission bias mode setting amplification means and the reception bias mode setting amplification means in accordance with the estimated values of the side noise power and the near-end side noise power, and the insertion loss. A quantity control means for estimating the instantaneous power of the input signal to the transmission side loss insertion means; and a second instantaneous power estimation section for estimating the instantaneous power of the input signal to the reception side loss insertion means. Instant Depending on the gain of the system that returns from the input point to the transmission side loss insertion means to the input point to the reception side loss insertion means through the transmission side loss insertion means and the wraparound on the line side. A line feedback gain multiplication unit having a value determined in accordance with a coefficient, and an input point to the reception-side loss insertion means to the reception-side loss insertion means and a wraparound on the acoustic side to the transmission-side loss insertion means An output signal obtained by inputting the output signal of the acoustic feedback gain multiplier having a value determined according to the gain of the path to the input point and the second instantaneous power estimator to the acoustic feedback gain multiplier And a first comparator for comparing the magnitude relationship between the output signal of the first instantaneous power estimation unit and the output signal obtained by inputting the output signal of the first instantaneous power estimation unit to the line feedback gain multiplier The magnitude of the output signal of the second instantaneous power estimator A second comparator for comparing the relationship, a transmission signal voice section detector for detecting a voice section of the transmission signal, a received signal voice section detector for detecting a voice section of the received signal, and a first comparator And determining the call state based on the comparison results of the second comparator and the detection results of the transmission signal voice section detection unit and the reception signal voice section detection unit, and inserting the transmission side loss insertion means and the reception side loss insertion means. An insertion loss amount distribution processing unit for controlling a loss amount, wherein the bias mode control means detects the magnitude relationship between the estimated values of the noise power on the far end side and the near end side and the detection of the received signal speech section detection unit Based on the result, the gain of the neutral mode for making the gains of the transmission bias mode setting amplifier and the amplification unit for setting the reception bias mode and the gain of the amplification unit for setting the transmission bias mode are set to the reception bias mode. Gain of amplification means Voice switching apparatus characterized by setting either of the transmission unbalance mode in which remote increases. 前記偏重モード制御手段は、前記遠端側雑音パワーの推定値が前記近端側雑音パワーの推定値と所定の第２の雑音パワー比係数との積以上の値となり且つ前記受話信号音声区間検出部によって音声区間が検出されない状態の継続時間が第２の所定時間以上となったときに送話偏重モード設定用増幅手段の利得を受話偏重モード設定用増幅手段の利得よりも増大させて送話偏重モードに設定することを特徴とする請求項３記載の音声切換装置。  The bias mode control means is configured such that the far-end side noise power estimate is equal to or greater than a product of the near-end side noise power estimate and a predetermined second noise power ratio coefficient, and the received signal speech section detection When the duration of the state in which the speech section is not detected by the unit becomes equal to or longer than the second predetermined time, the gain of the transmission bias mode setting amplification means is increased more than the gain of the reception bias mode setting amplification means. 4. The voice switching device according to claim 3, wherein the voice switching device is set to a bias mode. 前記送話側又は受話側の音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記しきい値並びに前記所定値を外部から設定可能としたことを特徴とする請求項１〜４の何れかに記載の音声切換装置。  The speech section detecting unit on the transmitting side or the receiving side includes a noise power estimating unit that estimates a noise power included in an input signal to the loss insertion means on the transmitting side or the receiving side, and the first or second instantaneous When the power estimate value is compared with a predetermined threshold value and the product of the noise power estimate value and the predetermined value, and the instantaneous power estimate value is greater than the threshold value and the instantaneous power estimate value is greater than the product And a voice / non-speech determination unit that determines that the input signal to the loss insertion means is a speech signal and determines that it is a non-speech signal when it is not large, and the threshold value and the predetermined value can be set from outside The voice switching device according to any one of claims 1 to 4, wherein 前記送話側又は受話側の音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記瞬時パワー推定部は立ち上がりが急峻で立ち下がりが緩やかな特性を有するフィルタからなり、前記雑音パワー推定部は立ち上がりが緩やかで立ち下がりが急峻な特性を有するフィルタからなり、該フィルタの前記特性を決定するパラメータを外部から設定可能としたことを特徴とする請求項１〜４の何れかに記載の音声切換装置。  The speech section detecting unit on the transmitting side or the receiving side includes a noise power estimating unit that estimates a noise power included in an input signal to the loss insertion means on the transmitting side or the receiving side, and the first or second instantaneous When the power estimate value is compared with a predetermined threshold value and the product of the noise power estimate value and the predetermined value, and the instantaneous power estimate value is greater than the threshold value and the instantaneous power estimate value is greater than the product And a speech / non-speech determination unit that determines that the input signal to the loss insertion means is a speech signal and determines that it is a non-speech signal when it is not large. The instantaneous power estimation unit has a sharp rise and a slow fall. The noise power estimation unit is composed of a filter having a slow rise and a steep fall characteristic, and determines the characteristic of the filter. Voice switching apparatus according to any one of claims 1 to 4, characterized in that a settable parameters from the outside. 前記しきい値及び所定値又は前記パラメータを前記送話側又は受話側の音声区間検出部に対して外部から個別に設定可能としたことを特徴とする請求項５又は６記載の音声切換装置。  7. The voice switching device according to claim 5, wherein the threshold value and the predetermined value or the parameter can be individually set from the outside to the voice section detecting unit on the transmitting side or the receiving side. 前記しきい値及び所定値又は前記パラメータを前記他の通話端末の種類に応じて各々異なる値に設定可能としたことを特徴とする請求項５又は６記載の音声切換装置。  7. The voice switching device according to claim 5, wherein the threshold value and the predetermined value or the parameter can be set to different values depending on the type of the other call terminal. 前記送話側又は受話側の音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記近端側雑音パワー推定手段を前記送話側音声区間検出部を構成する前記雑音パワー推定部で兼用するとともに、前記遠端側雑音パワー推定手段を前記受話側音声区間検出部を構成する前記雑音パワー推定部で兼用したことを特徴とする請求項１〜８の何れかに記載の音声切換装置。  The speech section detecting unit on the transmitting side or the receiving side includes a noise power estimating unit that estimates a noise power included in an input signal to the loss insertion means on the transmitting side or the receiving side, and the first or second instantaneous When the power estimate value is compared with a predetermined threshold value and the product of the noise power estimate value and the predetermined value, and the instantaneous power estimate value is greater than the threshold value and the instantaneous power estimate value is greater than the product And a voice / non-speech determination unit that determines that the input signal to the loss insertion means is a speech signal and determines that the signal is a non-speech signal when it is not large. The noise power estimation unit constituting the section detection unit is also used, and the far-end side noise power estimation unit is also used by the noise power estimation unit constituting the reception side voice section detection unit. Voice switching apparatus according to any one of claims 1 to 8. 前記偏重モード制御手段は、前記継続時間が前記第１の所定時間未満であっても、前記近端側雑音パワーの推定値が前記遠端側雑音パワーの推定値と前記第１の雑音パワー比係数よりも大きい所定の第３の雑音パワー比係数との積以上となる状態が前記第１の所定時間よりも長い所定の第３の所定時間以上継続したときには受話偏重モードに設定することを特徴とする請求項２記載の音声切換装置。  Even if the duration mode is less than the first predetermined time, the bias mode control means is configured such that the estimated value of the near-end side noise power is equal to the estimated value of the far-end side noise power and the first noise power ratio. When the state that is equal to or greater than the product of a predetermined third noise power ratio coefficient larger than the coefficient continues for a predetermined third predetermined time longer than the first predetermined time, the reception bias mode is set. The voice switching device according to claim 2. 前記偏重モード制御手段は、前記継続時間が前記第２の所定時間未満であっても、前記遠端側雑音パワーの推定値が前記近端側雑音パワーの推定値と前記第２の雑音パワー比係数よりも大きい所定の第４の雑音パワー比係数との積以上となる状態が前記第２の所定時間よりも長い所定の第４の所定時間以上継続したときには送話偏重モードに設定することを特徴とする請求項４記載の音声切換装置。  Even if the duration mode is less than the second predetermined time, the bias mode control means is configured such that the far-end side noise power estimated value is equal to the near-end side noise power estimated value and the second noise power ratio. When the state that is equal to or greater than the product of the predetermined fourth noise power ratio coefficient larger than the coefficient continues for a predetermined fourth predetermined time longer than the second predetermined time, the transmission bias mode is set. The voice switching device according to claim 4, wherein: 前記受話側損失挿入手段の出力点から近端側の音響エコー経路を介して前記送話側損失挿入手段の入力点へ帰還する経路の音響側帰還利得を推定する音響側帰還利得推定手段を備え、前記偏重モード制御手段は、音響側帰還利得の推定値が所定の条件を満たさなければ送話偏重モードに移行しないことを特徴とする請求項３記載の音声切換装置。Acoustic side feedback gain estimation means for estimating an acoustic side feedback gain of a path returning from an output point of the reception side loss insertion means to an input point of the transmission side loss insertion means via an acoustic echo path on the near end side; 4. The voice switching device according to claim 3 , wherein the bias mode control means does not shift to the transmission bias mode unless the estimated value of the acoustic feedback gain satisfies a predetermined condition. 前記偏重モード制御手段は、前記遠端側雑音パワーの推定値が前記近端側雑音パワーの推定値と前記第２の雑音パワー比係数との積以上であり、前記受話信号音声区間検出部の検出結果が非音声区間であり、且つ前記音響側帰還利得の推定値が所定のしきい値未満である状態が所定時間以上継続したときに送話偏重モードに設定することを特徴とする請求項１２記載の音声切換装置。  The bias mode control means is configured such that the estimated value of the far-end side noise power is not less than the product of the estimated value of the near-end side noise power and the second noise power ratio coefficient, The transmission bias mode is set when a detection result is a non-speech interval and a state where the estimated value of the acoustic feedback gain is less than a predetermined threshold continues for a predetermined time or more. 12. The voice switching device according to 12. 前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たさなければ受話偏重モードに移行しないことを特徴とする請求項１記載の音声切換装置。Line-side feedback gain estimation means for estimating a line-side feedback gain of a path that returns from the output point of the transmission-side loss insertion means to the input point of the reception-side loss insertion means via a far-end line echo path; the unbalance mode control means, the speech switching apparatus according to claim 1, wherein the estimated value of the line-side feedback gain is characterized in that it does not shift to the reception unbalance mode to meet a predetermined condition. 前記偏重モード制御手段は、前記近端側雑音パワーの推定値が前記遠端側雑音パワーの推定値と前記第１の雑音パワー比係数との積以上であり、前記送話信号音声区間検出部の検出結果が非音声区間であり、且つ前記回線側帰還利得の推定値が所定のしきい値未満である状態が所定時間以上継続したときに受話偏重モードに設定することを特徴とする請求項１４記載の音声切換装置。  The bias mode control means is configured such that the estimated value of the near-end side noise power is not less than the product of the estimated value of the far-end side noise power and the first noise power ratio coefficient, and the transmission signal speech section detecting unit The reception bias mode is set when the detection result is a non-voice interval and the estimated value of the line-side feedback gain is less than a predetermined threshold for a predetermined time or longer. 14. The voice switching device according to 14. 前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たさなければ受話偏重モードに移行せず、前記送話側及び受話側の各音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記送話側及び受話側の各音声区間検出部は、前記音響側並びに回線側の各帰還利得推定手段の推定値に応じて前記しきい値及び所定値が変更されることを特徴とする請求項１２記載の音声切換装置。  Line-side feedback gain estimation means for estimating a line-side feedback gain of a path that returns from the output point of the transmission-side loss insertion means to the input point of the reception-side loss insertion means via a far-end line echo path; The deviation mode control means does not shift to the reception deviation mode if the estimated value of the line-side feedback gain does not satisfy a predetermined condition, and the voice section detection units on the transmission side and the reception side A noise power estimator for estimating a noise power included in an input signal to the loss insertion means on the receiver side, the first or second instantaneous power estimated value as a predetermined threshold value, and the noise power estimated value and a predetermined value And when the instantaneous power estimate value is larger than the threshold value and the instantaneous power estimate value is larger than the product, the input signal to the loss insertion means is determined as a voice signal and greatly increased. A speech / non-speech determination unit that determines that a non-speech signal is detected, and the speech section detection units on the transmission side and the reception side are estimated values of the feedback gain estimation means on the acoustic side and the line side. 13. The voice switching device according to claim 12, wherein the threshold value and the predetermined value are changed according to the frequency. 前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たす場合には受話偏重モードに設定し、前記送話側及び受話側の各音声区間検出部は、送話側又は受話側の損失挿入手段への入力信号に含まれる雑音パワーを推定する雑音パワー推定部と、前記第１又は第２の瞬時パワー推定値を所定のしきい値並びに前記雑音パワー推定値と所定値の積とそれぞれ比較するとともに前記瞬時パワー推定値が前記しきい値より大きく且つ前記瞬時パワー推定値が前記積よりも大きいときに損失挿入手段への入力信号を音声信号と判定するとともに大きくないときに非音声信号と判定する音声／非音声判定部とを具備し、前記瞬時パワー推定部は立ち上がりが急峻で立ち下がりが緩やかな特性を有するフィルタからなり、前記雑音パワー推定部は立ち上がりが緩やかで立ち下がりが急峻な特性を有するフィルタからなり、前記送話側及び受話側の各音声区間検出部は、前記音響側並びに回線側の各帰還利得推定手段の推定値に応じて前記特性を決定するパラメータが変更されることを特徴とする請求項１２記載の音声切換装置。  Line-side feedback gain estimation means for estimating a line-side feedback gain of a path that returns from the output point of the transmission-side loss insertion means to the input point of the reception-side loss insertion means via a far-end line echo path; The deviation mode control means sets the reception deviation mode when the estimated value of the line-side feedback gain satisfies a predetermined condition, and each voice section detection unit on the transmission side and the reception side A noise power estimator for estimating a noise power included in an input signal to the loss insertion means on the receiver side, the first or second instantaneous power estimated value as a predetermined threshold value, and the noise power estimated value and a predetermined value And when the instantaneous power estimation value is larger than the threshold value and the instantaneous power estimation value is larger than the product, the input signal to the loss insertion means is determined to be a voice signal and becomes larger. A voice / non-speech determination unit that sometimes determines a non-speech signal, wherein the instantaneous power estimation unit is composed of a filter having a steep rise and a slow fall characteristic, and the noise power estimation unit has a slow rise The speech section detectors on the transmitting side and the receiving side determine the characteristics according to the estimated values of the feedback gain estimating means on the acoustic side and the line side. The voice switching device according to claim 12, wherein a parameter to be changed is changed. 前記しきい値及び所定値と前記パラメータを前記送話側並びに受話側の各音声区間検出部に対して外部から個別に設定可能としたことを特徴とする請求項１７記載の音声切換装置。  18. The voice switching device according to claim 17, wherein the threshold value, the predetermined value, and the parameter can be individually set from the outside for each voice section detecting unit on the transmitting side and the receiving side. 前記しきい値及び所定値と前記パラメータを前記他の通話端末の種類に応じて各々異なる値に設定可能としたことを特徴とする請求項１７記載の音声切換装置。  18. The voice switching device according to claim 17, wherein the threshold value, the predetermined value, and the parameter can be set to different values according to types of the other call terminals. 前記送話側損失挿入手段の出力点から遠端側の回線エコー経路を介して前記受話側損失挿入手段の入力点へ帰還する経路の回線側帰還利得を推定する回線側帰還利得推定手段を備え、前記偏重モード制御手段は、回線側帰還利得の推定値が所定の条件を満たさなければ受話偏重モードに移行せず、前記近端側雑音パワーの推定値と前記遠端側雑音パワーの推定値との差と、前記音響側並びに回線側の帰還利得推定値を参照して前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の利得を適応的に更新することを特徴とする請求項１２記載の音声切換装置。  Line-side feedback gain estimation means for estimating a line-side feedback gain of a path that returns from the output point of the transmission-side loss insertion means to the input point of the reception-side loss insertion means via a far-end line echo path; The deviation mode control means does not shift to the reception deviation mode if the estimated value of the line side feedback gain does not satisfy a predetermined condition, and the estimated value of the near end side noise power and the estimated value of the far end side noise power. And the gain of the transmission bias mode setting amplification means and the reception bias mode setting amplification means are adaptively updated with reference to the difference between the above and the feedback gain estimation values on the acoustic side and the line side. The voice switching device according to claim 12. 前記偏重モード制御手段により適応的に更新される前記送話偏重モード設定用増幅手段並びに受話偏重モード設定用増幅手段の利得に上限値を設定したことを特徴とする請求項２０記載の音声切換装置。  21. The voice switching device according to claim 20, wherein an upper limit value is set for the gain of the transmission bias mode setting amplification means and the reception bias mode setting amplification means that are adaptively updated by the bias mode control means. .
JP2003048260A 2003-02-25 2003-02-25 Audio switching device Expired - Fee Related JP3903928B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2003048260A JP3903928B2 (en) 2003-02-25 2003-02-25 Audio switching device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2003048260A JP3903928B2 (en) 2003-02-25 2003-02-25 Audio switching device

Publications (2)

Publication Number Publication Date
JP2004260491A JP2004260491A (en) 2004-09-16
JP3903928B2 true JP3903928B2 (en) 2007-04-11

Family

ID=33114254

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2003048260A Expired - Fee Related JP3903928B2 (en) 2003-02-25 2003-02-25 Audio switching device

Country Status (1)

Country Link
JP (1) JP3903928B2 (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4811039B2 (en) * 2006-02-06 2011-11-09 パナソニック電工株式会社 Audio switching device

Also Published As

Publication number Publication date
JP2004260491A (en) 2004-09-16

Similar Documents

Publication Publication Date Title
US7881927B1 (en) Adaptive sidetone and adaptive voice activity detect (VAD) threshold for speech processing
JP5086769B2 (en) Loudspeaker
JP4069595B2 (en) Loudspeaker
WO2006007516A2 (en) Limiting maximum gain in duplex communications devices including telephone sets
JP3492315B2 (en) Echo canceller with automatic volume adjustment
JP2001285513A (en) Digital cordless telephone with speakerphone incorporated in remote handset
JP3903928B2 (en) Audio switching device
JP2013187810A (en) Call signal processing apparatus
JP4131175B2 (en) Audio switching device
JP3941581B2 (en) Loudspeaker
JP4003739B2 (en) Loudspeaker
JP4380688B2 (en) Telephone device
CN103037121B (en) Squeaking elimination method and squeaking elimination system and hand-free phone
JP4811039B2 (en) Audio switching device
JP3709739B2 (en) Audio switching device
JP3941580B2 (en) Loudspeaker
JP4900185B2 (en) Loudspeaker
JP4003738B2 (en) Loudspeaker
JP4458002B2 (en) Loudspeaker
JP4333524B2 (en) Loudspeaker
JP3903933B2 (en) Telephone device
JP3726565B2 (en) Audio switching device
JP4211581B2 (en) Loudspeaker
JP4900184B2 (en) Loudspeaker
JP2012105217A (en) Handsfree telephone device, handsfree call method, and handsfree call program

Legal Events

Date Code Title Description
A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20050915

A977 Report on retrieval

Free format text: JAPANESE INTERMEDIATE CODE: A971007

Effective date: 20060911

A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20060919

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20061120

TRDD Decision of grant or rejection written
A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

Effective date: 20061219

A61 First payment of annual fees (during grant procedure)

Free format text: JAPANESE INTERMEDIATE CODE: A61

Effective date: 20070101

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20100119

Year of fee payment: 3

S533 Written request for registration of change of name

Free format text: JAPANESE INTERMEDIATE CODE: R313533

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20100119

Year of fee payment: 3

R350 Written notification of registration of transfer

Free format text: JAPANESE INTERMEDIATE CODE: R350

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20100119

Year of fee payment: 3

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20110119

Year of fee payment: 4

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20120119

Year of fee payment: 5

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20120119

Year of fee payment: 5

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20130119

Year of fee payment: 6

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20130119

Year of fee payment: 6

LAPS Cancellation because of no payment of annual fees