CN1432997A - Phonetic recognition method - Google Patents

Phonetic recognition method Download PDF

Info

Publication number
CN1432997A
CN1432997A CN02100044A CN02100044A CN1432997A CN 1432997 A CN1432997 A CN 1432997A CN 02100044 A CN02100044 A CN 02100044A CN 02100044 A CN02100044 A CN 02100044A CN 1432997 A CN1432997 A CN 1432997A
Authority
CN
China
Prior art keywords
voice
input
database
speech
sound bite
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
Application number
CN02100044A
Other languages
Chinese (zh)
Inventor
屈家伟
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.)
BITE LIANCHUANG ELECTRONIC (BEIJING) Co Ltd
Original Assignee
BITE LIANCHUANG ELECTRONIC (BEIJING) Co 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 BITE LIANCHUANG ELECTRONIC (BEIJING) Co Ltd filed Critical BITE LIANCHUANG ELECTRONIC (BEIJING) Co Ltd
Priority to CN02100044A priority Critical patent/CN1432997A/en
Publication of CN1432997A publication Critical patent/CN1432997A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Electrically Operated Instructional Devices (AREA)

Abstract

The present invention discloses a phonetic recognition method includes the specific steps of: recording standard phonetic data in phonetic data base; indexing and selecting standard voice segment to be displayed, prompted or broadcast; outputting standard voice segment, inputting new voice segment; recognizing through comparing the input voice signal and that in the data base; judging and feedback displaying. The present invention utilizes the recognizable and un-recognizable voice simultaneously in accurate judgment.

Description

A kind of audio recognition method
Technical field
The present invention relates to a kind of audio recognition method, this method is mainly used in voice activated control, man-machine conversation field.
Background technology
Existing speech recognition technology mainly is an identifiability of utilizing voice, and speech recognition system is carried out operation accordingly according to the voice signal of external world's input.When speech recognition system can be discerned extraneous voice signal, speech recognition system was just carried out corresponding operation; When speech recognition system can't be discerned extraneous voice signal, speech recognition system or ignore was perhaps carried out wrong understanding, the operation of execution error.Promptly have only the correct voice signal of input, speech recognition system could correctly be discerned, and for the input of mistake, speech recognition system can not be discerned, and can't judge.
With the product that popular in the market major part help learns a language, be example as radio cassette player, learning machine, language repeater etc., the recognition methods of its voice is: play segment standard recording earlier; After the learner hears out, repeat to read through; At this moment, play again after learner's sound recording can being entered; Compare by learner oneself, judge.The shortcoming of this audio recognition method is to be carried out the identification judgement of voice by people's ear oneself, because everyone standard is different, oneself listens unavoidable some subjectivity of pronunciation of oneself, often is not easy to find wrong or non-type place, the shortage objectivity.
Summary of the invention
In view of the foregoing, the purpose of this invention is to provide a kind of can be objective, judge exactly, the method for recognizing voice.
For achieving the above object, the present invention takes following design proposal: a kind of audio recognition method, and concrete steps are as follows:
The first step: the received pronunciation data are logged in the speech database;
Second step: index, the sound bite of selection standard in the received pronunciation database;
The 3rd step: outputting standard sound bite;
The 4th step: import new voice signal;
Convert new voice signal to speech data and be stored in the speech data buffer by recording device, analog-digital commutator;
The 5th step: the voice signal of standard in the voice signal of identification, relatively input and the database;
By the voice recognition chip in the speech recognition device new voice signal and the received pronunciation in the speech database are analyzed, discern, compared;
The 6th step: judge whether correctly give clear and definite answer to new voice signal, and display result;
After the identification of voice recognition chip, comparison, judgement, if the voice of input are correct, then give affirmative answer, and restart next circulation, promptly repeated for first~the 6th step; Through judging, if the voice of input are incorrect, then give negative answer, and show mistake, re-enter, repeated for the 4th~the 6th step.
In the audio recognition method of the present invention the 5th, the 6th step is that the voice of standard in the new voice that will import by voice recognition chip and the speech database are analyzed, discerned, comparison.
Because the present invention adopts above technical scheme, promptly by speech recognition system Direct Recognition, comparison, judge whether new voice signal correct, and the judged result feedback is shown.Be identifiability and the nonidentifiability that the present invention has utilized voice simultaneously, thereby the judgement of standard is more accurate, objective, just to new voice signal.
Description of drawings
The invention will be further described below in conjunction with accompanying drawing.
Fig. 1 is a speech recognition system theory diagram of the present invention;
Fig. 2 is an audio recognition method process flow diagram of the present invention;
Fig. 3 is a voice recognition chip control program process flow diagram.
Embodiment
As shown in Figure 1, speech recognition system of the present invention mainly is made up of recording device 1, analog-digital commutator 2, speech data buffer 3, speech database 4, speech recognition device 5 and output unit 6; Wherein, store the speech recognition program that embodies audio recognition method of the present invention in the speech recognition device 5 in advance.Speech recognition device 5 of the present invention adopts the speech recognition family chip.
As shown in Figure 2, audio recognition method of the present invention is:
The first step: the received pronunciation data are logged in the speech database;
Second step: index, the sound bite of selection standard in the received pronunciation database;
The 3rd step: outputting standard sound bite;
Can play-over this received pronunciation fragment, also can show this received pronunciation fragment, perhaps point out the content of this received pronunciation fragment by other modes by display.
The 4th step: import new sound bite:
Convert new sound bite to speech data and be stored in the speech data buffer by recording device, analog-digital commutator;
The 5th step: the voice signal of standard in the voice signal of identification, relatively input and the database;
By the voice recognition chip in the speech recognition device voice and the received pronunciation in the database of input are analyzed, discern, compared.This identification, relatively be the function that voice recognition chip itself just has, only need in addition that corresponding programmed control gets final product, as shown in Figure 3.
The 6th step: judge whether correctly give clear and definite answer to new voice signal, and display result;
After the identification of voice recognition chip, comparison, judgement, if the voice of input are correct, give affirmative answer, and restart next circulation, promptly repeated for first~the 6th step; Through judging, if the voice of input are incorrect, give negative answer, and clearly inform the phonetic entry mistake, re-enter, repeated for the 4th~the 6th step.
This method has been utilized the identifiability and the nonidentifiability of voice simultaneously.When speech recognition system can be discerned external sound, think that the voice of input are correct; When system can not discern, think the voice mistake of input and to give clear and definite informing, the judged result feedback is shown.
This method can specifically be applied in the interactive learning methods such as language playback device, learning machine, its specific implementation process is: add the speech recognition system that includes audio recognition method of the present invention in learning machine, during use, the learner is every to say a word or expression, speech recognition system is just carried out objective, just differentiation and analysis to learner's voice, can identify, think that the learner carries a tune; Can not identify, think learner's mispronounce, and the result is fed back to the learner.

Claims (3)

1. audio recognition method, concrete steps are as follows:
The first step: the received pronunciation data are logged in the speech database;
Second step: index, the sound bite of selection standard in the received pronunciation database;
The 3rd step: outputting standard sound bite;
The 4th step: import new sound bite;
The 5th step: the voice signal of standard in the voice signal of identification, relatively input and the database;
The 6th step: judge, give clear and definite answer to two kinds of voice are whether identical, and show feedback.
2. a kind of audio recognition method according to claim 1 is characterized in that: described the 5th, the 6th step is meant by voice recognition chip analyzes the voice and the received pronunciation in the speech database of input, discern, compare.
3. a kind of audio recognition method according to claim 1 and 2 is characterized in that: described audio recognition method concrete steps are as follows:
The first step: the received pronunciation data are logged in the speech database;
Second step: index, the sound bite of selection standard in the received pronunciation database;
The 3rd step: outputting standard sound bite;
The 4th step: import new sound bite;
The learner converts new sound bite to speech data and is stored in the speech data buffer by recording device, analog-digital commutator;
The 5th step: the voice signal of standard in the voice signal of identification, relatively input and the database;
By speech recognition device, voice recognition chip the voice and the received pronunciation in the speech database of input are analyzed, discern, compared;
The 6th step: judge whether correctly give clear and definite answer, and show feedback to new voice signal;
After the identification of voice recognition chip, comparison, judgement, if the voice of input are correct, give affirmative answer, and restart next circulation, promptly repeated for first~the 6th step; Through judging, if the voice of input are incorrect, give negative answer, clearly inform input voice mistake, re-enter, repeated for the 4th~the 6th step.
CN02100044A 2002-01-09 2002-01-09 Phonetic recognition method Pending CN1432997A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN02100044A CN1432997A (en) 2002-01-09 2002-01-09 Phonetic recognition method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN02100044A CN1432997A (en) 2002-01-09 2002-01-09 Phonetic recognition method

Publications (1)

Publication Number Publication Date
CN1432997A true CN1432997A (en) 2003-07-30

Family

ID=27626963

Family Applications (1)

Application Number Title Priority Date Filing Date
CN02100044A Pending CN1432997A (en) 2002-01-09 2002-01-09 Phonetic recognition method

Country Status (1)

Country Link
CN (1) CN1432997A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102723080A (en) * 2012-06-25 2012-10-10 惠州市德赛西威汽车电子有限公司 Voice recognition test system and voice recognition test method
CN105427387A (en) * 2015-11-09 2016-03-23 上海语知义信息技术有限公司 System and method for controlling ticket vending machine by voice
CN107305769A (en) * 2016-04-20 2017-10-31 斑马网络技术有限公司 Voice interaction processing method, device, equipment and operating system
CN109036402A (en) * 2018-07-18 2018-12-18 深圳市本牛科技有限责任公司 Digital speech VOD system and its operating method and the device for using the system
CN109767773A (en) * 2019-03-26 2019-05-17 北京百度网讯科技有限公司 Information output method and device based on interactive voice terminal

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102723080A (en) * 2012-06-25 2012-10-10 惠州市德赛西威汽车电子有限公司 Voice recognition test system and voice recognition test method
CN105427387A (en) * 2015-11-09 2016-03-23 上海语知义信息技术有限公司 System and method for controlling ticket vending machine by voice
CN107305769A (en) * 2016-04-20 2017-10-31 斑马网络技术有限公司 Voice interaction processing method, device, equipment and operating system
CN109036402A (en) * 2018-07-18 2018-12-18 深圳市本牛科技有限责任公司 Digital speech VOD system and its operating method and the device for using the system
CN109767773A (en) * 2019-03-26 2019-05-17 北京百度网讯科技有限公司 Information output method and device based on interactive voice terminal

Similar Documents

Publication Publication Date Title
CN104252864B (en) Real-time voice analysis method and system
US6332122B1 (en) Transcription system for multiple speakers, using and establishing identification
CN1145141C (en) Method and device for improving accuracy of speech recognition
CN1235188C (en) Correcting a text recognized by speech recognition through comparison of phonetic sequences in the recognized text with a phonetic transcription of a manually input correction word
EP2506252A3 (en) Topic specific models for text formatting and speech recognition
CN106558252B (en) Spoken language practice method and device realized by computer
CN111711834B (en) Recorded broadcast interactive course generation method and device, storage medium and terminal
CN1908863A (en) Two language mixed input method and mobile phone with dictionary function
CN111415537A (en) Symbol-labeling-based word listening system for primary and secondary school students
CN105575402A (en) Network teaching real time voice analysis method
CN101079189A (en) Chinese pronunciation correction listening-writing study method and system
CN1458645A (en) Voice identification equipment and voice identification program
CN1432997A (en) Phonetic recognition method
CN1424665A (en) Device and operation for dictation test and automatic
CN111710203A (en) English pronunciation correction system based on big data
CN113205717A (en) Deep learning-based oral English training method
CN1267805C (en) User's interface, system and method for automatically marking phonetic symbol to correct pronunciation
CN204926573U (en) Intelligent robot of auxiliary exercise mandarin
CN1228760C (en) System and method for searching word according to speech
CN112102670B (en) Teaching auxiliary system based on voice control and working method thereof
US20050158696A1 (en) [interactive computer-assisted language learning method and system thereof]
CN1204826A (en) Method for deaf-dumb voice learning dialogue and pronunciation synchronous feedback device
CN111356022A (en) Video file processing method based on voice recognition
CN1763800A (en) Cartoon type language learning system and its cartoon producing method
CN1074552C (en) User's action recording device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication