JPH0384600A - Voice recognition device - Google Patents

Voice recognition device

Info

Publication number
JPH0384600A
JPH0384600A JP1222738A JP22273889A JPH0384600A JP H0384600 A JPH0384600 A JP H0384600A JP 1222738 A JP1222738 A JP 1222738A JP 22273889 A JP22273889 A JP 22273889A JP H0384600 A JPH0384600 A JP H0384600A
Authority
JP
Japan
Prior art keywords
model
voice
speech
input
similar
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.)
Granted
Application number
JP1222738A
Other languages
Japanese (ja)
Other versions
JP2966002B2 (en
Inventor
Yoshiharu Abe
芳春 阿部
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.)
Mitsubishi Electric Corp
Original Assignee
Mitsubishi Electric Corp
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 Mitsubishi Electric Corp filed Critical Mitsubishi Electric Corp
Priority to JP1222738A priority Critical patent/JP2966002B2/en
Publication of JPH0384600A publication Critical patent/JPH0384600A/en
Application granted granted Critical
Publication of JP2966002B2 publication Critical patent/JP2966002B2/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Abstract

PURPOSE:To enable high-accuracy voice recognition even when a large vocabulary including many similar words is an object by performing rerecognition by using a recognizing means suitable to the recognition of the similar words when high-order candidates include similar words. CONSTITUTION:Tolerance to an input feature vector sequence is calculated as to each model generated in consideration of a context and a primary selection part 5 selects the model having the highest tolerance. A similar phoneme sequence detection part 6 detects similar phoneme sequences among candidates for an input voice selected by the primary selection part 5 and an emphasized model generation part 7 generates the model of a voice by emphasizing the part of a phoneme different among those similar phoneme sequences. Then a reselection part 8 selects the emphasized voice model having the highest tolerance to the input voice among the generated emphasized voice models and outputs the phoneme sequence as the source for the generation as a candidate for the input voice. Consequently, even when a large vocabulary including many similar words is an object of recognition, high recognition accuracy is obtained.

Description

【発明の詳細な説明】 〔産業上の利用分野〕 本発明は、単語音声や文音声の認識を行う音声認識装置
に関し、特に大語賃を認識対象とする音声認識装置の改
良に関するものである。 〔従来の技術〕[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a speech recognition device that recognizes word speech and sentence speech, and particularly relates to an improvement of a speech recognition device that recognizes large word sounds. . [Conventional technology]

Claims (1)

【特許請求の範囲】[Claims] 　音声の発声内容を記述した記号の系列から該記号系列
の表す音声のモデルを該記号の文脈あるいは該記号に対
応する入力音声の部分の文脈に依存して生成するモデル
生成手段と、該モデル生成手段で生成された音声モデル
の中で入力音声に対する尤度の高い音声モデルを選択し
その生成の源となった記号の系列を入力音声の候補とし
て出力する一次選択手段とを備えた音声認識装置におい
て、上記一次選択手段で選択された入力音声の候補の中
で類似の記号系列を検出する類似記号系列検出手段と、
該類似記号系列検出手段で類似の記号系列が検出された
場合にこれらの類似の記号系列の間の相違する記号の部
分を強調した音声のモデルを生成する強調モデル生成手
段と、該強調モデル生成手段で生成された強調音声モデ
ルの中で入力音声に対する尤度の高い強調音声モデルを
選択しその生成の源となった記号系列を入力音声の候補
として出力する再選択手段とを設けたことを特徴とする
音声認識装置。a model generation means for generating a model of the speech represented by the symbol series from a series of symbols describing the content of speech utterances depending on the context of the symbol or the context of the part of the input speech corresponding to the symbol; A speech recognition device comprising primary selection means for selecting a speech model with a high likelihood for the input speech from among the speech models generated by the means and outputting the sequence of symbols from which the model was generated as a candidate for the input speech. similar symbol sequence detection means for detecting similar symbol sequences among the input speech candidates selected by the primary selection means;
Emphasis model generation means for generating a speech model that emphasizes different symbol parts between the similar symbol series when similar symbol series are detected by the similar symbol series detection means; and the emphasis model generation means. Re-selection means is provided for selecting an emphasized voice model having a high likelihood for the input voice from among the emphasized voice models generated by the means, and outputting the symbol sequence from which the model was generated as a candidate for the input voice. Characteristic voice recognition device.
JP1222738A 1989-08-29 1989-08-29 Voice recognition device Expired - Fee Related JP2966002B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP1222738A JP2966002B2 (en) 1989-08-29 1989-08-29 Voice recognition device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP1222738A JP2966002B2 (en) 1989-08-29 1989-08-29 Voice recognition device

Publications (2)

Publication Number Publication Date
JPH0384600A true JPH0384600A (en) 1991-04-10
JP2966002B2 JP2966002B2 (en) 1999-10-25

Family

ID=16787129

Family Applications (1)

Application Number Title Priority Date Filing Date
JP1222738A Expired - Fee Related JP2966002B2 (en) 1989-08-29 1989-08-29 Voice recognition device

Country Status (1)

Country Link
JP (1) JP2966002B2 (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016010184A1 (en) * 2014-07-16 2016-01-21 이인지 Pet hair cutting device
US9645584B2 (en) 2014-09-17 2017-05-09 Honeywell International Inc. Gas valve with electronic health monitoring
US9657946B2 (en) 2012-09-15 2017-05-23 Honeywell International Inc. Burner control system
US9683674B2 (en) 2013-10-29 2017-06-20 Honeywell Technologies Sarl Regulating device

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS60121499A (en) * 1983-12-05 1985-06-28 富士通株式会社 Voice collation system
JPS63148299A (en) * 1986-12-12 1988-06-21 株式会社日立製作所 Word voice recognition equipment
JPH01167896A (en) * 1987-12-24 1989-07-03 Toshiba Corp Voice input device

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS60121499A (en) * 1983-12-05 1985-06-28 富士通株式会社 Voice collation system
JPS63148299A (en) * 1986-12-12 1988-06-21 株式会社日立製作所 Word voice recognition equipment
JPH01167896A (en) * 1987-12-24 1989-07-03 Toshiba Corp Voice input device

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9657946B2 (en) 2012-09-15 2017-05-23 Honeywell International Inc. Burner control system
US9683674B2 (en) 2013-10-29 2017-06-20 Honeywell Technologies Sarl Regulating device
WO2016010184A1 (en) * 2014-07-16 2016-01-21 이인지 Pet hair cutting device
US9645584B2 (en) 2014-09-17 2017-05-09 Honeywell International Inc. Gas valve with electronic health monitoring

Also Published As

Publication number Publication date
JP2966002B2 (en) 1999-10-25

Similar Documents

Publication Publication Date Title
US5787230A (en) System and method of intelligent Mandarin speech input for Chinese computers
JP4536323B2 (en) Speech-speech generation system and method
US20180137109A1 (en) Methodology for automatic multilingual speech recognition
CN106486121B (en) Voice optimization method and device applied to intelligent robot
JP2815579B2 (en) Word candidate reduction device in speech recognition
US10803858B2 (en) Speech recognition apparatus, speech recognition method, and computer program product
Kurzekar et al. Continuous speech recognition system: A review
Maqsood et al. A comparative study of classifier based mispronunciation detection system for confusing
JPH06110494A (en) Pronounciation learning device
JP6599914B2 (en) Speech recognition apparatus, speech recognition method and program
JPH0384600A (en) Voice recognition device
Prukkanon et al. F0 contour approximation model for a one-stream tonal word recognition system
KR20060067107A (en) Continuous speech recognition apparatus using articulatory model and method thereof
JP3039453B2 (en) Voice recognition device
KR102405547B1 (en) Pronunciation evaluation system based on deep learning
JPH10232693A (en) Voice recognition device
JP2001188556A (en) Method and device for voice recognition
JP2001005483A (en) Word voice recognizing method and word voice recognition device
KR20050058949A (en) Prosodic phrasing method for korean texts
JP2001013983A (en) Speech recognition apparatus using speech synthesis and speech recognition method
El Méliani et al. Specific language modelling for new-word detection in continuous-speech recognition
JPH03287200A (en) Voice recognition device
JP2004309654A (en) Speech recognition apparatus
JPH04291399A (en) Voice recognizing method
JPS62166400A (en) Voice wordprocessor

Legal Events

Date Code Title Description
LAPS Cancellation because of no payment of annual fees