JPS63158595A - Voice recognition equipment - Google Patents
Voice recognition equipmentInfo
- Publication number
- JPS63158595A JPS63158595A JP61307417A JP30741786A JPS63158595A JP S63158595 A JPS63158595 A JP S63158595A JP 61307417 A JP61307417 A JP 61307417A JP 30741786 A JP30741786 A JP 30741786A JP S63158595 A JPS63158595 A JP S63158595A
- Authority
- JP
- Japan
- Prior art keywords
- recognition
- section
- speech
- voice
- mode
- 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
- 238000001514 detection method Methods 0.000 claims description 15
- 238000000034 method Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
- 230000001755 vocal effect Effects 0.000 description 1
Abstract
(57)【要約】本公報は電子出願前の出願データであるた
め要約のデータは記録されません。(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.
Description
【発明の詳細な説明】 （産業上の利用分野） 本発明は、音声認識装置の改良に関するものである。[Detailed description of the invention] (Industrial application field) The present invention relates to improvements in speech recognition devices.
（従来の技術）
近年、音声認識技術と半導体技術の進歩により、予め限
定された複数個の孤立単語を認識することのできる音声
認識装置が開発されている。その応用として、特願昭５
６−１６９７１号、￥？ＰＪ昭５６−１６９７２号。(Prior Art) In recent years, with advances in speech recognition technology and semiconductor technology, speech recognition devices that can recognize a plurality of predefined isolated words have been developed. As an application of this,
No. 6-16971, ¥? PJ No. 56-16972.
特願昭５６−１６９７３号、特願昭５７−３３５８７号
等の明細書に、操作者の発声した命令音声に対応した動
作を被制御系に指令する信号を出力する音声認識による
制御装置が、実例をあげて示されている。音声認識の手
法に関しては、１９８５年９月、東海大学出版会より出
版された「デジタル音声処理」の第１４９頁より第１９
２頁に記載されている。Japanese Patent Application No. 56-16973, Japanese Patent Application No. 57-33587, etc. disclose a control device using voice recognition that outputs a signal instructing a controlled system to perform an operation corresponding to a command voice uttered by an operator. Illustrated with examples. Regarding speech recognition methods, see pages 149 to 19 of "Digital Speech Processing" published by Tokai University Press in September 1985.
It is described on page 2.
（発明が解決しようとする問題点）
上記のような従来の音声認識による制御装置は、背景雑
音が大きくて認識が行えない場合、他人と会話する場合
等で、認識動作を一時中断する必要がある時には、マイ
クもしくは音声認識装置のスイッチを手操作で切るか、
音声入力で認識装置を一時中断させる命令を発生しなけ
ればならなかった。前者の方法では、両手が自由に使え
るという音声認識装置の利点が生かされていないし、後
者の方法では、雑音が大きくて認識が行えない場合は認
識部πの中断ができないという問題があった。(Problems to be Solved by the Invention) Conventional voice recognition control devices as described above require temporary interruption of recognition operation when background noise is too loud to perform recognition, or when talking with another person. In some cases, you may manually turn off the microphone or voice recognition device, or
A command to temporarily suspend the recognition device had to be generated using voice input. The former method does not take advantage of the speech recognition device's ability to use both hands freely, and the latter method has the problem that the recognition section π cannot be interrupted when recognition cannot be performed due to large noise.
また、特開昭５９−２２１００号に示されているように
、背景雑音が大きい場合は、認識装置が強制的に認識動
作を行わないというものもあるが、繰作者の意思で認識
動作を中断・再開できないといった欠点があった。In addition, as shown in Japanese Patent Application Laid-open No. 59-22100, there is a system in which the recognition device is forced not to perform the recognition operation when the background noise is large, but the recognition operation is interrupted at the user's will.・There was a drawback that it could not be restarted.
本発明の目的は、上記問題点を解決し、雑音下でも操作
者の意思で手軽に認諾動作の中断および再開ができる音
声認識装置を提供することにある。SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems and to provide a speech recognition device that allows the operator to easily interrupt and restart the approval operation even under noisy conditions.
（問題点を解決するための手段）
本発明の音声認識装置は、入力された音声を分析して特
徴量を計算する分析部と、前記音声の存在する区間であ
る音声区間を前記特徴量に基づき検出する音声検出部と
、前記音声検出部で検出された音声を認識する認識部と
前記音声検出部で検出された前記音声区間内の前記音声
の定常区間と検出する音声定常区間検出部と、前記認識
部に認識動作を行わせる認識モードと前記認識部に認識
動作を行わせない認識中断モードとを前記音声の定常区
間が一定時間以上の場合に切り替えるモード切り替え部
とを有する。(Means for Solving the Problems) The speech recognition device of the present invention includes an analysis unit that analyzes input speech and calculates a feature amount, and a speech section in which the speech exists to calculate the feature amount. a recognition unit that recognizes the voice detected by the voice detection unit; and a voice steady interval detection unit that detects the voice within the voice interval detected by the voice detection unit; and a mode switching unit that switches between a recognition mode in which the recognition unit performs a recognition operation and a recognition interruption mode in which the recognition unit does not perform a recognition operation when the steady section of the voice is longer than a certain time.
（作用） 本発明の詳細な説明する。(effect) The present invention will be described in detail.
操作者は、認識装置を一時中断させたい場合、音声の母
音をある一定時開発声する。母音を引き伸ばして発声し
た場合、その音響的特徴は定常となり、ある程度の雑音
下でも検出は容易に行うことができる。認識装置は、定
常的な音声を検出したら、その継続時間を測定する。そ
れが、ある一定時間以下である場合は、今までどおりの
認識動作を行い、ある一定時間以上である場合は、認識
動作を一時中断するものと判断して、次に同様な一定時
間以上継続する定常的な音声を検出するまで、入力され
た音声に対して認識動作を行わない。If the operator wants to temporarily interrupt the recognition device, he/she speaks a vocal vowel for a certain period of time. When a vowel is elongated and uttered, its acoustic characteristics become stationary and can be easily detected even under a certain amount of noise. When the recognition device detects a steady sound, it measures its duration. If it is less than a certain period of time, the recognition operation continues as before, and if it is longer than a certain period of time, the recognition operation is judged to be temporarily interrupted and then continued for a similar period of time. No recognition operation is performed on the input voice until a steady voice is detected.
操作者が、認識動作を再開させたい場合は、再び、先と
同様の音声の母音をある一定時間発声して、認諾装置に
一定時間以上継続枕する定常的な音声を検出させればよ
い。If the operator wishes to restart the recognition operation, he/she may utter the same vowel sound as before for a certain period of time again and have the recognition device detect a steady sound that continues for a certain period of time or more.
以上が、本発明の作用である。The above is the operation of the present invention.
（実施例）
以下、本発明の実施例について図面を参照して説明する
。第１図は、本発明の一実施例の音声認識装置を示すブ
ロック図である。(Example) Hereinafter, an example of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a speech recognition device according to an embodiment of the present invention.
本実施例の音声認識装置は、入力された音声の認識を行
う認識モードと、音声が入力されても認識を行わない認
識中断モードとの２つのモードで動作が異なり、モード
切り替え部６で、モードが切り替わる。The speech recognition device of this embodiment operates in two modes: a recognition mode in which input speech is recognized, and a recognition interruption mode in which speech is not recognized even if speech is input. The mode will change.
マイクロホン１から入力された音声信号は、音声分析部
２において、たとえばｎ願昭５２−１４１１２０５号明
細書及びその第３図に示された如き周波数分析器によっ
て、音声分析がなされ、ベクトルの時系列ａｌ （ｔ）
、ａ２ （ｔ）、・・・ａｉ（ｔ）、・・・ａｎ （ｔ
）＜ａｔ （ｔ）は、時刻しにおけるベクトルの第１番
目の要素）に変換されて、音声検出部３に送られる。以
下、このベクトル時系列を入カバターンと称する。音声
検出部３では入カバターンのエネルギーを監視すること
により、音声区間の検出を行い、音声が検出されたらそ
の区間の入カバターンを逐次音声定常区間検出部４に送
る。The audio signal input from the microphone 1 is subjected to audio analysis in the audio analysis section 2 using a frequency analyzer such as that shown in the specification of Japanese Patent No. 52-1411205 and its FIG. al(t)
, a2 (t), ... ai (t), ... an (t
)<at (t) is converted into (the first element of the vector at the time) and sent to the voice detection section 3. Hereinafter, this vector time series will be referred to as an input pattern. The voice detecting section 3 detects a voice section by monitoring the energy of the input cover turn, and when a voice is detected, sequentially sends the input cover turn of that section to the voice steady section detecting section 4.
なお、音声検出に関しては、１９７９年、共立出版株式
会社より出版された「音声認識」の第６８頁がら第７０
頁に記載されている。音声定常区間検出部４では、入カ
バターンの各フレーム間の差分Ｄ（ｔ）＝Σｌａｉ （
ｔ）−ａｉ （ｔ−１） ｌ■
を計算し、その値がある閾値Ｄｈ以下となる区間が一定
時間Ｔｈ１以上続いた場合、定常区間始端検出信号Ｓｌ
をタイマ一部うに送る。その後、フレーム間の差分があ
る閾値Ｄｈより大きくなった場合、もしくは、入カバタ
ーンが終了した場合、定常区間終端検出信号Ｓ２をタイ
マ一部５に送る。Regarding voice detection, see pages 68 to 70 of ``Speech Recognition'' published by Kyoritsu Shuppan Co., Ltd. in 1979.
It is written on the page. In the voice steady section detection unit 4, the difference D(t)=Σlai (
t)-ai (t-1) l■ is calculated, and if the interval in which the value is below a certain threshold Dh continues for a certain period of time Th1 or more, the steady interval start detection signal Sl
is sent to part of the timer. Thereafter, when the difference between frames becomes larger than a certain threshold value Dh, or when the input cover turn ends, a steady section end detection signal S2 is sent to the timer part 5.
タイマ一部５では、Ｓｌから８２までの時間１゛を計測
して、モード切り習え部６に送る。The timer section 5 measures the time 1'' from Sl to 82 and sends it to the mode learning section 6.
モード切り替え部６では、時間Ｔがある閾値Ｔｈ２以上
であった場合のみ、モードの変更を行い、認識モード開
始信号Ｓ３、もしくは認識中断モード開始信号Ｓ４を、
その都度、認識制御部７に送る。認識制御部７は、認識
モード開始信号Ｓ３を受信すると、認識部８に認識モー
ドの動作（音声検出部３から、逐次入カバターンを受は
取り、認識を行って認識結果を出力する）を行わせるよ
う指令する。また、認識制御部７は認識中断モード開始
信号Ｓ４を受信すると、認識部８に認識中断モードの動
作（何も行わない）を行わせるよう指令する。認識部８
は、音声認識の分野では周知のＤＰマツチテンを使って
マルチテンプレート法を用いたもの、たとえば、特開昭
４９−７９４３９号の実施例に記載されている如くに構
成される。認識部８については、本発明の主旨ではない
ので、詳細は省く。The mode switching unit 6 changes the mode only when the time T is equal to or greater than a certain threshold Th2, and sends the recognition mode start signal S3 or the recognition interruption mode start signal S4.
Each time, it is sent to the recognition control unit 7. Upon receiving the recognition mode start signal S3, the recognition control unit 7 causes the recognition unit 8 to operate in the recognition mode (receives and receives sequential input cover turns from the voice detection unit 3, performs recognition, and outputs the recognition result). command to do so. When the recognition control unit 7 receives the recognition interruption mode start signal S4, it instructs the recognition unit 8 to perform the recognition interruption mode operation (do nothing). Recognition unit 8
is constructed using a multi-template method using DP Matsuchiten, which is well known in the field of speech recognition, for example, as described in the embodiment of JP-A-49-79439. The details of the recognition unit 8 will be omitted since it is not the gist of the present invention.
第２図は以上の動作の概要を示すフローチャートである
。FIG. 2 is a flowchart showing an overview of the above operation.
なお、認諾部８には、他にも、特開昭５９−９１５００
号に記載されている如く構成される音声認識の分野では
周知のヒドウン・マルコフ・モデルを用いたもの、１９
８５年９月、東海大学出版会から出版された「デジタル
音声処理」の第１８７頁から第１８８頁に記載されてい
る識別関数方式を用いたもの等が適用できる。In addition, the approval section 8 also has Japanese Patent Application Laid-open No. 59-91500.
In the field of speech recognition, the one using the well-known hidden Markov model is configured as described in No. 19.
The method using the discriminant function method described on pages 187 to 188 of "Digital Speech Processing" published by Tokai University Press in September 1985 can be applied.
（発明の効果）
以上述べたとおり、本発明によれば、雑音下でも操作者
の意思で手軽に認識動作の中断および再開ができる音声
認識装置を提供することができる。(Effects of the Invention) As described above, according to the present invention, it is possible to provide a speech recognition device in which the recognition operation can be easily interrupted and restarted at the operator's will even under noise.
第１図は本発明の一実施例の音声認識装置を示すブロッ
ク図、第２図は本実施例の音声認識装置における動作を
示すフローチャートである。
１・・・マイクロホン、２・・・分析部、３・・・音声
検出部、４・・・定常区間検出部、５・・・タイマ一部
、６・・・モード切り替え部、７・・・認識制御部、８
・・・認識部。FIG. 1 is a block diagram showing a speech recognition apparatus according to an embodiment of the present invention, and FIG. 2 is a flowchart showing the operation of the speech recognition apparatus according to this embodiment. DESCRIPTION OF SYMBOLS 1...Microphone, 2...Analysis section, 3...Speech detection section, 4...Steady interval detection section, 5...Timer part, 6...Mode switching section, 7... Recognition control unit, 8
...Recognition part.
Claims (1)
前記音声の存在する区間である音声区間を前記特徴量に
基づき検出する音声検出部と、前記音声検出部で検出さ
れた前記音声区間内の前記音声を認識する認識部とを有
する音声認識装置において、前記音声検出部で検出され
た前記音声区間内の前記音声の定常区間を検出する音声
定常区間検出部と、前記認識部に認識動作を行わせる認
識モードと前記認識部に認識動作を行わせない認識中断
モードとを前記音声の定常区間が一定時間以上の場合に
切り替えるモード切り替え部とを有することを特徴とす
る音声認識装置。an analysis unit that analyzes the input voice and calculates feature quantities;
A speech recognition device comprising: a speech detection section that detects a speech section in which the speech exists based on the feature amount; and a recognition section that recognizes the speech within the speech section detected by the speech detection section. , a voice steady section detection section that detects a steady section of the voice within the voice section detected by the voice detection section; a recognition mode that causes the recognition section to perform a recognition operation; and a recognition mode that causes the recognition section to perform a recognition operation. a mode switching section that switches between a recognition interruption mode and a recognition interruption mode when the steady section of the voice is longer than a certain period of time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP61307417A JPS63158595A (en) | 1986-12-22 | 1986-12-22 | Voice recognition equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP61307417A JPS63158595A (en) | 1986-12-22 | 1986-12-22 | Voice recognition equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPS63158595A true JPS63158595A (en) | 1988-07-01 |
Family
ID=17968805
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP61307417A Pending JPS63158595A (en) | 1986-12-22 | 1986-12-22 | Voice recognition equipment |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPS63158595A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08115096A (en) * | 1994-10-14 | 1996-05-07 | Sanyo Electric Co Ltd | Voice processor |
| JP2016099479A (en) * | 2014-11-20 | 2016-05-30 | アイシン・エィ・ダブリュ株式会社 | Voice control system, voice control method, and voice control program |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS6136238A (en) * | 1984-07-12 | 1986-02-20 | ローン‐プーラン・サント | Manufacture of unsaturated compounds alpha_substituted in beta_position of two electron_withdrawing groups |
-
1986
- 1986-12-22 JP JP61307417A patent/JPS63158595A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS6136238A (en) * | 1984-07-12 | 1986-02-20 | ローン‐プーラン・サント | Manufacture of unsaturated compounds alpha_substituted in beta_position of two electron_withdrawing groups |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08115096A (en) * | 1994-10-14 | 1996-05-07 | Sanyo Electric Co Ltd | Voice processor |
| JP2016099479A (en) * | 2014-11-20 | 2016-05-30 | アイシン・エィ・ダブリュ株式会社 | Voice control system, voice control method, and voice control program |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP3674990B2 (en) | Speech recognition dialogue apparatus and speech recognition dialogue processing method | |
| US9775113B2 (en) | Voice wakeup detecting device with digital microphone and associated method | |
| CN110265036A (en) | Voice awakening method, system, electronic equipment and computer readable storage medium | |
| US11308946B2 (en) | Methods and apparatus for ASR with embedded noise reduction | |
| CN110867197A (en) | Method and equipment for interrupting voice robot in real time in voice interaction process | |
| JP3211398B2 (en) | Speech detection device for video conference | |
| CN110689887B (en) | Audio verification method and device, storage medium and electronic equipment | |
| JP3553828B2 (en) | Voice storage and playback method and voice storage and playback device | |
| KR20050049207A (en) | Dialogue-type continuous speech recognition system and using it endpoint detection method of speech | |
| JPS63158595A (en) | Voice recognition equipment | |
| JPS62150295A (en) | Voice recognition | |
| JPH0950288A (en) | Device and method for recognizing voice | |
| JP3846500B2 (en) | Speech recognition dialogue apparatus and speech recognition dialogue processing method | |
| JP2019132997A (en) | Voice processing device, method and program | |
| JP3940895B2 (en) | Speech recognition apparatus and method | |
| JPS59137999A (en) | Voice recognition equipment | |
| JP4143487B2 (en) | Time-series information control system and method, and time-series information control program | |
| JPH03114100A (en) | Voice section detecting device | |
| JPH04163497A (en) | Voice section detecting method | |
| JP2000039900A (en) | Speech interaction device with self-diagnosis function | |
| JP2737109B2 (en) | Voice section detection method | |
| JPH02103599A (en) | Voice recognizing device | |
| JP2017201348A (en) | Voice interactive device, method for controlling voice interactive device, and control program | |
| JPS6370298A (en) | Double consonant recognition equipment | |
| JP2000089799A (en) | Voice recognition system and method and recording medium stored with software for voice recognietion |