JPS59225441A - Voice input device - Google Patents
Voice input deviceInfo
- Publication number
- JPS59225441A JPS59225441A JP58100517A JP10051783A JPS59225441A JP S59225441 A JPS59225441 A JP S59225441A JP 58100517 A JP58100517 A JP 58100517A JP 10051783 A JP10051783 A JP 10051783A JP S59225441 A JPS59225441 A JP S59225441A
- Authority
- JP
- Japan
- Prior art keywords
- voice
- section
- display
- signal
- input device
- 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.)
- Pending
Links
Abstract
Description
【発明の詳細な説明】 １１立！ 本発明は、音声入力装置に関する。[Detailed description of the invention] 11 standing! The present invention relates to a voice input device.
良末且遣
従来の音声入力装置は、入力音声の大きさを表わすレベ
ルメータを具備しており、オペレータはこれを見ながら
発声し、声の大きさを調整していた。しかし、この方法
では、実際の作業として、データな読んで発声する場合
に、目は対象物と認識結果を確認する表示器を見る方に
専念し、レベルメータを見ることはほとんしない、従っ
て１発声する声の大きさは経時変化し、大きくなったり
小さくなったりし、装置の電気回路のダイナミックレン
ジを越えて波形が歪んだり、入力が小さいため特徴抽出
の精度が悪くなって安定した動作を保証することができ
ない等の欠点があった。Conventional voice input devices are equipped with a level meter that indicates the volume of the input voice, and the operator speaks while looking at the level meter to adjust the volume of the voice. However, with this method, when reading and uttering data as part of the actual work, the eyes are focused on looking at the target object and the display that confirms the recognition results, and hardly ever look at the level meter. The volume of the voice that is uttered changes over time, becoming louder or softer, and the waveform may become distorted by exceeding the dynamic range of the device's electrical circuit, or the accuracy of feature extraction may deteriorate due to the small input, making it difficult to operate stably. There were drawbacks such as the inability to provide guarantees.
−−１
本発明は、上述のごとき実情に鑑みてなされたもので、
特に、音声入力装置に入力される音声の大きさを監視し
、音声の大きさに異常があれば、オペレータに声を大き
くすべきか−小さくすべきかを知らせ、常に適切な音量
で入力し安定した音声認識精度を確保することを目的と
してなされたものである。--1 The present invention was made in view of the above-mentioned circumstances, and
In particular, we monitor the volume of the voice input to the voice input device, and if there is an abnormality in the volume of the voice, we inform the operator whether the voice should be raised or lowered, and always input at an appropriate volume to maintain a stable level. This was done with the purpose of ensuring speech recognition accuracy.
梃−一一痕
本発明の構成について、以下、実施例に基づいて説明す
る。The structure of the present invention will be described below based on examples.
本発明は、音声入力装置に入力される音声信号のエネル
ギー（平均電力）及び最大振幅によって入力される音声
の大きさが適切か否かを調べて表示し、又は、オペレー
タに知らせることを特徴とする。また、エネルギー、又
は、最大振幅に異常が検出された場合に、その異常の内
容をオペレータに知らせるか、異常が極端な場合には認
識計算を中断してリセットし、その異常内容をオペレー
タに知らせて再入力を促すか、或いは他の手段例えばキ
ー人力を指示することを特徴とするものである。The present invention is characterized by checking and displaying or notifying the operator whether or not the volume of the input voice is appropriate based on the energy (average power) and maximum amplitude of the voice signal input to the voice input device. do. In addition, if an abnormality is detected in energy or maximum amplitude, the operator is notified of the abnormality, or if the abnormality is extreme, the recognition calculation is interrupted and reset, and the operator is notified of the abnormality. This feature is characterized by prompting the user to re-enter the key, or instructing other means such as key manual input.
第１図は、本発明の一実施例を説明するための全体回路
図で、音響−電気変換器ｌ（例えばマイクロホン）で収
音された音声信号は増幅器２によって増幅され、自動利
得制御回路（例えばＡＧＣ）３で子音及び母音エネルギ
ーの差を極力小さくして音声入力の振幅変動を吸収し、
音声切出し部４で音声信号のエネルギー情報やその他の
情報から音声区間を検出し、認識部５で音声信号を分析
し、特徴量を抽出・して音声の認識を行なう。同時に増
幅器２の出力は、音量検出部６の一方の入力端に入力さ
れ異常振幅の検出をする。また、自動利得制御回路３の
出力も音量検出部６のもう一方の入与端に入力される。FIG. 1 is an overall circuit diagram for explaining one embodiment of the present invention, in which an audio signal picked up by an acoustic-electrical converter l (for example, a microphone) is amplified by an amplifier 2, and an automatic gain control circuit ( For example, AGC) 3 minimizes the difference between consonant and vowel energy to absorb amplitude fluctuations in voice input,
A speech cutting section 4 detects a speech section from the energy information and other information of the speech signal, and a recognition section 5 analyzes the speech signal, extracts feature quantities, and performs speech recognition. At the same time, the output of the amplifier 2 is input to one input terminal of the volume detection section 6 to detect abnormal amplitude. Further, the output of the automatic gain control circuit 3 is also input to the other input terminal of the volume detection section 6.
音量検出部６では最大振幅とエネルギーの最小値が音声
区間において所定のレベル内にあるかどうかを測定し、
表示部９に信号を送り、ディスプレー１０に表示する。The volume detection unit 6 measures whether the maximum amplitude and the minimum value of energy are within a predetermined level in the voice section,
A signal is sent to the display unit 9 and displayed on the display 10.
もし、所定のレベル内にない場合には認識部５に信号を
送って認識計算の中断を命令してリセットし、一方、表
示部９に信号を送ってディスプレーｌＯに表示すると共
に音声応答部７に信号を送ってその旨をレシーバ８に出
力する。If it is not within the predetermined level, a signal is sent to the recognition unit 5 to instruct it to interrupt the recognition calculation and reset, while a signal is sent to the display unit 9 to display it on the display lO, and at the same time, the voice response unit 7 and outputs a signal to that effect to the receiver 8.
第２図は、第１図に示した音量検出部６の動作説明をす
るための詳細電気回路図で、図中、第１図と同様の作用
をする部分には第１図の場合と同一の参照番号を付しで
ある。音響−電気変換器１の音声信号は増幅器２で増幅
され、自動利得制御回路３及び比較器１１に供給される
。比較器１１は端子１２に加えられた電圧と比較し、音
声信号が端子１２の定電圧より大きくなると出力信号を
出し、２人力アンド回路１３の一方に入力する。FIG. 2 is a detailed electrical circuit diagram for explaining the operation of the volume detection section 6 shown in FIG. with reference numbers. The audio signal from the acousto-electrical converter 1 is amplified by an amplifier 2 and supplied to an automatic gain control circuit 3 and a comparator 11. The comparator 11 compares the voltage applied to the terminal 12, and when the audio signal becomes larger than the constant voltage of the terminal 12, it outputs an output signal and inputs it into one of the two-manual AND circuits 13.
２人力アン・ド回路１３のもう一方の入力は音声切出し
部より端子２６に送られて来た音声区間信号であり、出
力端子２２より表示部９または音声応答部７に信号が送
られ、異常に大きい振幅の信号が入力されたことをオペ
レータに知らせる。なお、端子１２の定電圧は自動利得
制御回路以降の最大許″容入力電圧によって決定される
。また、増幅器２から自動利得制御回路３を経た信号は
、検波回路１４及び積分回路１５を通してエネルギー（
平均電力）に変換され、次いで比較器１６によって端子
１７の電圧と比較される。比較器１６は反転型であるの
でエネルギーが端子１７の電圧より小さいときに出力信
号を出し、２人力アンド回路１８の一方に入力されると
共に２人力ＮＯＲ回路１９の一方に入力される。２人力
アンド回路１８のもう一方の入力は音声区間信号であり
、出力端子２５より表示部９または音声応答部７に信号
が送られ、音声信号が小さすぎる旨をオペレータに知ら
せる。なお、端子１７の定電圧は自動利得制御回路以降
のダイナミックレンジ及び音声の切出し精度、認識精度
を勘案して決定される。また、２人力ＮＯＲ回路１９の
もう一方の入力端には、比較器１１の出方が加えられ音
声信号が所定のレベル内にあるとき出力信号を出し、２
人力ＡＮＤ回路２０の一方に人力される。２人力ＡＮＤ
’回路２０のもう一方の入力端には音声区間
信号が加えられ、出力端子２３から表示部９に信号が送
られ、音声信号が適切なレベルにあることをティスプレ
ーｌＯに示す。また、ＡＮＤ回路１３及び１８の出力は
ＯＲ回路２１に加えられ、出力端子２４より認識部５に
信号が送られ、認識計算を中断してリセットする。The other input of the two-man power AND/DO circuit 13 is the audio section signal sent from the audio cutout section to the terminal 26, and the signal is sent from the output terminal 22 to the display section 9 or the audio response section 7 to detect an abnormality. informs the operator that a large amplitude signal has been input. Note that the constant voltage at the terminal 12 is determined by the maximum allowable input voltage after the automatic gain control circuit. Furthermore, the signal that has passed from the amplifier 2 to the automatic gain control circuit 3 is converted into energy (
average power) and then compared with the voltage at terminal 17 by comparator 16. Since the comparator 16 is of an inverting type, it outputs an output signal when the energy is smaller than the voltage at the terminal 17, which is input to one side of the two-way AND circuit 18 and one side of the two-way NOR circuit 19. The other input of the two-manpower AND circuit 18 is a voice section signal, and the signal is sent from the output terminal 25 to the display section 9 or the voice response section 7 to inform the operator that the voice signal is too small. Note that the constant voltage of the terminal 17 is determined by taking into consideration the dynamic range after the automatic gain control circuit, the audio extraction accuracy, and the recognition accuracy. In addition, the output of the comparator 11 is added to the other input terminal of the two-man power NOR circuit 19, and when the audio signal is within a predetermined level, it outputs an output signal.
The human power is applied to one side of the human power AND circuit 20. 2 person AND
A voice interval signal is applied to the other input of the circuit 20, and a signal is sent from the output terminal 23 to the display 9, indicating on the display lO that the voice signal is at the appropriate level. Further, the outputs of the AND circuits 13 and 18 are applied to the OR circuit 21, and a signal is sent from the output terminal 24 to the recognition unit 5 to interrupt and reset the recognition calculation.
丸−−Ｊ
以北の説明から明らかなように１本発明によると音声入
力装置に入力される音量を監視し、その結果をオペレー
タに表示又は音声でフィードバックすることによって認
識に最適な音量を保持して安定な認識率を確保すること
ができる。また、音声の音量が所定のレベルに入らない
場合も含めて異常入力時に、認識計算を中断して誤認識
を防ぐことができる等の利点がある。As is clear from the above description, according to the present invention, the volume input to the voice input device is monitored, and the result is displayed or voiced back to the operator to maintain the optimal volume for recognition. It is possible to ensure a stable recognition rate. Another advantage is that recognition calculations can be interrupted to prevent erroneous recognition when an abnormal input is made, including when the volume of the voice does not fall within a predetermined level.
第１図は、本発明の一実施例を説明するための電気回路
図、第２図は、第１図に示した音量検出部６の動作説明
をするための詳細電気回路図である。
１・・−音響−電気変換器、２・・・増幅器、３・・・
自動利得制御回路、４・・音声切出し部、５・・・認識
部、６・・・音量検出部、７・・・音声応答部、８・・
・レシーバ、９・・・表示部、ｌＯ・・・ディスプレー
。FIG. 1 is an electric circuit diagram for explaining one embodiment of the present invention, and FIG. 2 is a detailed electric circuit diagram for explaining the operation of the volume detecting section 6 shown in FIG. 1...-acoustic-electrical converter, 2... amplifier, 3...
automatic gain control circuit, 4... audio extraction unit, 5... recognition unit, 6... volume detection unit, 7... audio response unit, 8...
・Receiver, 9...display section, lO...display.
Claims (3)
し、該音声の電気信号を分析して種々の特徴量を抽出し
て音声を認識する音声入力装置において、音声信号の振
幅の最大値又はエネルギーが所。 定のレベルに人っているか否かを検出する検出器、及び
、所定のレベル内に入っていない場合に、その旨をオペ
レータに知らせる表示器又は音響応答器を有することを
特徴とする音声入力装置。(1) In a voice input device that recognizes voice by converting voice into an electrical signal using an acoustic-electrical converter, analyzing the electrical signal of the voice, and extracting various feature quantities, the maximum value of the amplitude of the voice signal Or where the energy is. an audio input device characterized in that it has a detector for detecting whether or not a person is within a predetermined level; and an indicator or an acoustic transponder to notify the operator if the person is not within the predetermined level; Device.
ネルギーが小さすぎるかを前記表示器又は音声応答器に
よってオペレータに知らせるようにしたことを特徴とす
る特許請求の範囲第（１）項に記載の音声入力装置。(2) Claim (1) characterized in that the display or the voice response device informs the operator whether the amplitude of the voice signal is abnormally large or the energy is too small. The voice input device described.
場合に、音声認識ｍｔｉ：）認識計算を中断することを
特徴とする特許請求の範囲第（１）項または第（２）項
に記載の音声入力装置。(3) If the level of the voice does not fall within a predetermined level, the voice recognition mti:) recognition calculation is interrupted, as set forth in claim (1) or (2). voice input device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP58100517A JPS59225441A (en) | 1983-06-06 | 1983-06-06 | Voice input device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP58100517A JPS59225441A (en) | 1983-06-06 | 1983-06-06 | Voice input device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPS59225441A true JPS59225441A (en) | 1984-12-18 |
Family
ID=14276141
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP58100517A Pending JPS59225441A (en) | 1983-06-06 | 1983-06-06 | Voice input device |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPS59225441A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS60151745A (en) * | 1984-01-20 | 1985-08-09 | Hitachi Ltd | Voice information input device |
| JPS61221928A (en) * | 1985-03-28 | 1986-10-02 | Hitachi Ltd | Inputting device for voice information |
| JPS62297929A (en) * | 1986-06-13 | 1987-12-25 | インタ−ナショナル ビジネス マシ−ンズ コ−ポレ−ション | Document processing system |
| JPH04134398A (en) * | 1990-09-26 | 1992-05-08 | Matsushita Electric Ind Co Ltd | Voice recognizing device |
| WO2003052737A1 (en) * | 2001-12-17 | 2003-06-26 | Asahi Kasei Kabushiki Kaisha | Speech recognition method, remote controller, information terminal, telephone communication terminal and speech recognizer |
| CN106847306A (en) * | 2016-12-26 | 2017-06-13 | 华为技术有限公司 | The detection method and device of a kind of abnormal sound signal |
-
1983
- 1983-06-06 JP JP58100517A patent/JPS59225441A/en active Pending
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS60151745A (en) * | 1984-01-20 | 1985-08-09 | Hitachi Ltd | Voice information input device |
| JPS61221928A (en) * | 1985-03-28 | 1986-10-02 | Hitachi Ltd | Inputting device for voice information |
| JPH0215897B2 (en) * | 1985-03-28 | 1990-04-13 | Hitachi Ltd | |
| JPS62297929A (en) * | 1986-06-13 | 1987-12-25 | インタ−ナショナル ビジネス マシ−ンズ コ−ポレ−ション | Document processing system |
| JPH04134398A (en) * | 1990-09-26 | 1992-05-08 | Matsushita Electric Ind Co Ltd | Voice recognizing device |
| WO2003052737A1 (en) * | 2001-12-17 | 2003-06-26 | Asahi Kasei Kabushiki Kaisha | Speech recognition method, remote controller, information terminal, telephone communication terminal and speech recognizer |
| JP2009104156A (en) * | 2001-12-17 | 2009-05-14 | Asahi Kasei Homes Kk | Telephone communication terminal |
| CN106847306A (en) * | 2016-12-26 | 2017-06-13 | 华为技术有限公司 | The detection method and device of a kind of abnormal sound signal |
| CN106847306B (en) * | 2016-12-26 | 2020-01-17 | 华为技术有限公司 | Abnormal sound signal detection method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108172242B (en) | Improved Bluetooth intelligent cloud sound box voice interaction endpoint detection method | |
| US7167544B1 (en) | Telecommunication system with error messages corresponding to speech recognition errors | |
| CN109065064B (en) | Method for generating EQ curve, method for outputting audio and output equipment | |
| JPS59225441A (en) | Voice input device | |
| WO2016067644A1 (en) | Speech adjustment device | |
| CN106658324A (en) | Hearing aid with verification function and verification method | |
| JP3131226B2 (en) | Hearing aid with improved percentile predictor | |
| JPH0635497A (en) | Speech input device | |
| JP2807241B2 (en) | Voice recognition device | |
| JPH0627986A (en) | Equipment control system utilizing speech recognizing device | |
| JPS6147438B2 (en) | ||
| JPH02232697A (en) | Voice recognition device | |
| US6338036B1 (en) | Confirmation notification by apparatus using audio recognition as to the acceptability of an input sound | |
| JP3056048B2 (en) | Snoring detection device | |
| JPH02176796A (en) | Speech recognition device | |
| US11758337B2 (en) | Audio processing apparatus | |
| JPH02103599A (en) | Voice recognizing device | |
| KR100939684B1 (en) | Voice recorder with 3 microphone | |
| JPH04340598A (en) | Voice recognition device | |
| JP2008284151A (en) | Cough detector and cough detection program | |
| JP3632384B2 (en) | Hearing aids | |
| JPH039400A (en) | Voice recognizer | |
| JPS63163494A (en) | Intensity detector | |
| JPH04186399A (en) | Speech recognition device | |
| JP3474072B2 (en) | Voice recognition device and voice recognition method |