CN108304385A - A kind of speech recognition text error correction method and device - Google Patents

A kind of speech recognition text error correction method and device Download PDF

Info

Publication number
CN108304385A
CN108304385A CN201810135187.4A CN201810135187A CN108304385A CN 108304385 A CN108304385 A CN 108304385A CN 201810135187 A CN201810135187 A CN 201810135187A CN 108304385 A CN108304385 A CN 108304385A
Authority
CN
China
Prior art keywords
text
error correction
user
candidate
speech recognition
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
CN201810135187.4A
Other languages
Chinese (zh)
Inventor
叶伟
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.)
Shanghai Century Network Technology Co., Ltd.
Original Assignee
叶伟
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 叶伟 filed Critical 叶伟
Priority to CN201810135187.4A priority Critical patent/CN108304385A/en
Publication of CN108304385A publication Critical patent/CN108304385A/en
Priority to PCT/CN2019/070817 priority patent/WO2019153996A1/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G06F16/3344—Query execution using natural language analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/332—Query formulation
    • G06F16/3329—Natural language query formulation or dialogue systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
    • G06F16/68—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/683—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G06F16/685—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using automatically derived transcript of audio data, e.g. lyrics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/232—Orthographic correction, e.g. spell checking or vowelisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/237—Lexical tools
    • G06F40/242—Dictionaries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/26—Speech to text systems

Abstract

A kind of speech recognition text error correction method, this method include：Generate user-oriented dictionary library；It generates candidate and waits for corrected text set；Text collection after the candidate error correction of generation；And text collection after candidate error correction is screened, obtain text after error correction.The embodiment of the present invention constructs the judgment models of text after error correction, generates text after candidate error correction by establishing the technical characteristic of user's corpus；And establish user's sentence prediction model, using the model, text after the error correction of speech recognition is calculated, one of advantageous effect of acquisition is, in Intelligent housing field, further implementation is provided for the accuracy of the speech recognition controlled of intelligent appliance.

Description

A kind of speech recognition text error correction method and device
Technical field
The invention belongs to technical field of voice recognition, more particularly to a kind of speech recognition text error correction method and device.
Background technology
With popularizing for deep learning, obtained in computer vision, speech recognition, natural language processing etc. great It breaks through.By taking speech recognition as an example, speech recognition accuracy has reached 97% at present.The breakthrough of the above technology so that speech recognition Application field it is more and more wider.Since relative to other man-machine interaction modes, interactive voice more meets the daily habits of people, It is highly efficient.It is contemplated that speech recognition technology will be widely used in smart home, industrial production, communication, medical treatment, drive automatically The every field such as sail.
In actual speech interactive process, since each factors such as user pronunciation is nonstandard, noise influence, sound identifies error rate It is higher.And the prior art all concentrates on being promoted on speech recognition accuracy, but lacks the approach of error correction to recognition result.The above original Cause, extreme influence interactive voice product promotion.
Invention content
The embodiment provides a kind of speech recognition text error correction method and devices, it is therefore intended that solves voice and knows For the Error Correcting Problem of speech recognition text in other technology.
One of the embodiment of the present invention, a kind of speech recognition text error correction method, including,
User-oriented dictionary library generating method：Counting user family often uses language material, and is arranged.According to the language material text after arrangement This, carries out text participle, part-of-speech tagging, word frequency statistics, pinyin marking.
Candidate waits for corrected text generation method：First, it after being segmented to speech recognition text, remove stop words, checks Whether each word is in user-oriented dictionary library.If in user-oriented dictionary library, without error correction.If not making in user-oriented dictionary library Corrected text is waited for for candidate.
Document creation method after candidate error correction：First according to training sample, generation judges whether the word is text after error correction Judgment models, wherein using editing distance as mode input feature.Text after all candidate error correction is replaced and waits for error correction text This, to generate text after candidate error correction.
In terms of Intelligent housing, text screening technique after candidate error correction：By all user data originally counted, packet It includes the conduct of the parameters such as User ID, interactive voice text time, sensing data, family's electricity condition and is originally inputted parameter.After arranging User's language material of gained is intended to and entity is as output, builds user's sentence prediction model.Using the model, when by interactive voice The parameters such as time, sensing data, family's electricity condition input, to obtain the user view and entity that prediction obtains.It will predict institute Obtained user view and entity carries out similarity calculation with text after candidate error correction, after obtaining the highest candidate error correction of similarity Text is as text after unique error correction.
The embodiment of the present invention constructs the judgement mould of text after error correction by establishing the technical characteristic of user's corpus Type generates text after candidate error correction；And it establishes user's sentence prediction model and speech recognition is calculated using the model Text after error correction, one of advantageous effect of acquisition are, are the speech recognition controlled of intelligent appliance in Intelligent housing field Accuracy provides further implementation.
Description of the drawings
Detailed description below, above-mentioned and other mesh of exemplary embodiment of the invention are read by reference to attached drawing , feature and advantage will become prone to understand.In the accompanying drawings, if showing the present invention's by way of example rather than limitation Dry embodiment, wherein：
Fig. 1 is the flow chart of the speech recognition error correction method based on user's corpus in one embodiment of the invention.
Fig. 2 is that user-oriented dictionary library builds flow chart in one embodiment of the invention.
Fig. 3 is that text model flow figure after candidate error correction is built in one embodiment of the invention.
Specific implementation mode
According to one or more embodiments, as shown in Figure 1, being the flow signal of voice recognition result text error correction method Figure.Process step includes,
S11:Build user-oriented dictionary library.Each user version language material being collected into is subjected to cleaning arrangement, and segmented, gone Except stop words.It is one according to word frequency, part of speech, phonetic by word segmentation result completely to index, structure user-oriented dictionary library, and wherein each Phonetic decollator is divided.
S12:Quasi- candidate waits for that corrected text generates.User speech is interacted into text after removing stop words, obtains each point Word result.It checks whether each word occurs in user-oriented dictionary library, if in each user-oriented dictionary library occurring, waits entangling not as candidate Wrong text, it is on the contrary then wait for corrected text as candidate.When inspection, examined in user-oriented dictionary library one by one for each word after having divided It looks into, error correction is not necessarily to if word is in user-oriented dictionary library；If word is just used as not in user-oriented dictionary library and waits for corrected text.
S13:Text generation after candidate error correction.The corpus data being collected into before first, compiles and respectively waits correcting Text with correct after the text editing distance of text, initial editing distance, full pinyin editing distance, and will respectively edit above away from From being normalized as input feature vector, text generation model after candidate error correction is carried out.It calculates and respectively waits for corrected text and user The text editing distance of each word, initial editing distance, full pinyin editing distance in dictionary, and as mode input, Show which word is as candidate corrected text in user-oriented dictionary library, which is not as candidate corrected text.If only there are one wait Text after error correction is selected, then as result after text error correction；If there is text after multiple candidate error correction, then continue candidate Text screens after error correction.
S14:From being screened in text after each candidate error correction, text after most possible error correction is picked out.Involved in the step To using machine learning techniques, user's language material prediction model based on user environment is built.Interactive voice is generated according to user The dimensions such as time, place, sensing data, appliance data build user's language material prediction model.It includes mainly being intended to, being real that it, which is exported, Body target, control mode.Such as output：Home wiring control is intended to, entity is air-conditioning, control mode is heating.By text after each error correction Similarity calculation is carried out with model output result, chooses after the highest candidate error correction of similarity text as unique error correction hereinafter This.
According to one or more embodiments, as shown in Fig. 2, user-oriented dictionary library builds flow chart.User-oriented dictionary library generation side Method includes：Counting user family often uses language material, and is arranged.According to the language material text after arrangement, text participle, part of speech are carried out Mark, word frequency statistics, pinyin marking, structure user-oriented dictionary library.
According to one or more embodiments, as shown in figure 3, text judgment models flow diagram after the candidate error correction of structure.
S31 arranges raw tone and identifies text, and desensitizes；
S32 manually carries out error correction to raw tone identification text, and records text after each error correction；
S33 counts respectively text and each word editing distance in user-oriented dictionary library after error correction, this editing distance includes text editing Distance, first spelling editing distance, full pinyin editing distance, and each editing distance is normalized, the input as model；
S34, marks whether each word in dictionary is text after final correct.If so, 1 is labeled as, if not then marking It is 0；
S35, according to the above each mark sample, using correlation machine learning algorithm, generation judges whether after being candidate error correction The model of text.
According to one or more embodiments, a kind of speech recognition text error correction method device, which includes memory；With And it is coupled to the one or more processors of the memory, processor is configured as executing the finger being stored in the memory It enables, the processor executes following operation：
Generate user-oriented dictionary library；
It generates candidate and waits for corrected text set；
Text collection after the candidate error correction of generation；And
Text collection after candidate error correction is screened, text after error correction is obtained.
The memory includes computer-readable record/storage medium, such as random access memory (RAM), read-only storage Device (ROM), flash memories, CD, disk, solid-state disk etc..
It is worth noting that although foregoing teachings are by reference to several essences that detailed description of the preferred embodimentsthe present invention has been described creates God and principle, it should be appreciated that, the present invention is not limited to disclosed specific implementation mode, the division also unawareness to various aspects The feature that taste in these aspects cannot combine, this to divide the convenience merely to statement.The present invention is directed to cover appended power Included various modifications and equivalent arrangements in the spirit and scope that profit requires.

Claims (12)

1. a kind of speech recognition text error correction method, which is characterized in that this method includes：
Generate user-oriented dictionary library；
It generates candidate and waits for corrected text set；
Text collection after the candidate error correction of generation；And
Text collection after candidate error correction is screened, text after error correction is obtained.
2. speech recognition text error correction method as described in claim 1, which is characterized in that the step of generating user-oriented dictionary library is wrapped It includes：
User version language material is collected, and is arranged；
Stop words is segmented and is removed to the language material text after arrangement；
User-oriented dictionary library is built after carrying out part-of-speech tagging, word frequency statistics, pinyin marking to word segmentation result.
3. speech recognition text error correction method as described in claim 1, which is characterized in that generate candidate and wait for corrected text set Step includes：
After being segmented to speech recognition text, remove stop words, each word segmentation result is obtained；
Each participle is checked whether in user-oriented dictionary library, if occurring in user-oriented dictionary library, the speech recognition text is not made Corrected text is waited for for candidate, if not in user-oriented dictionary library, the speech recognition text waits for corrected text as candidate.
4. speech recognition text error correction method as described in claim 1, which is characterized in that text collection after the candidate error correction of generation Step includes：
According to training sample, generate judge each word in user-oriented dictionary library whether be text after candidate error correction judgment models, wherein Using editing distance as mode input feature；
Text after all candidate error correction is replaced and waits for corrected text, to generate text after candidate error correction.
5. speech recognition text error correction method as described in claim 1, which is characterized in that carried out to text collection after candidate error correction Include the step of text after acquisition error correction after screening：
User's language material is intended to as parameter is originally inputted and entity is as output, structure is used by the user data that statistics is obtained Family sentence prediction model；
Using user's sentence prediction model, the initial parameter including the interactive voice time of speech recognition text is inputted, to Obtain the user view and entity that prediction obtains；
It will predict that obtained user view and entity carry out similarity calculation with text after candidate error correction, and obtain similarity highest Candidate error correction after text as text after unique error correction.
6. speech recognition text error correction method as claimed in claim 4, which is characterized in that text judgment models after candidate error correction Construction step includes：
It arranges raw tone and identifies text；
Error correction manually is carried out to raw tone identification text, and records text after each error correction；
Respectively text and each word editing distance in user-oriented dictionary library after error correction, this editing distance include text editing distance, head to statistics Editing distance, full pinyin editing distance are spelled, and each editing distance is normalized, the input as model；
Whether each word is text after final correct in mark user-oriented dictionary library.If so, 1 is labeled as, if not being then labeled as 0；
According to each mark sample, using correlation machine learning algorithm, generate judge whether be text after candidate error correction model.
7. a kind of speech recognition text error correction method device, which is characterized in that the device includes memory；And
It is coupled to the one or more processors of the memory, processor is configured as executing and be stored in the memory Instruction, the processor execute following operation：
Generate user-oriented dictionary library；
It generates candidate and waits for corrected text set；
Text collection after the candidate error correction of generation；And
Text collection after candidate error correction is screened, text after error correction is obtained.
8. speech recognition text error correction device as claimed in claim 7, which is characterized in that the step of generating user-oriented dictionary library is wrapped It includes：
User version language material is collected, and is arranged；
Stop words is segmented and is removed to the language material text after arrangement；
User-oriented dictionary library is built after carrying out part-of-speech tagging, word frequency statistics, pinyin marking to word segmentation result.
9. speech recognition text error correction device as claimed in claim 7, which is characterized in that generate candidate and wait for corrected text set Step includes：
After being segmented to speech recognition text, remove stop words, each word segmentation result is obtained；
Each participle is checked whether in user-oriented dictionary library, if occurring in user-oriented dictionary library, the speech recognition text is not made Corrected text is waited for for candidate, if not in user-oriented dictionary library, the speech recognition text waits for corrected text as candidate.
10. speech recognition text error correction device as claimed in claim 7, which is characterized in that generate text collection after candidate error correction The step of include：
According to training sample, generate judge each word in user-oriented dictionary library whether be text after candidate error correction judgment models, wherein Using editing distance as mode input feature；
Text after all candidate error correction is replaced and waits for corrected text, to generate text after candidate error correction.
11. speech recognition text error correction device as claimed in claim 7, which is characterized in that text collection after candidate error correction into Include the step of text after acquisition error correction after row screening：
User's language material is intended to as parameter is originally inputted and entity is as output, structure is used by the user data that statistics is obtained Family sentence prediction model；
Using user's sentence prediction model, the initial parameter including the interactive voice time of speech recognition text is inputted, to Obtain the user view and entity that prediction obtains；
It will predict that obtained user view and entity carry out similarity calculation with text after candidate error correction, and obtain similarity highest Candidate error correction after text as text after unique error correction.
12. speech recognition text error correction device as claimed in claim 7, which is characterized in that text judgment models after candidate error correction Construction step include：
It arranges raw tone and identifies text；
Error correction manually is carried out to raw tone identification text, and records text after each error correction；
Respectively text and each word editing distance in user-oriented dictionary library after error correction, this editing distance include text editing distance, head to statistics Editing distance, full pinyin editing distance are spelled, and each editing distance is normalized, the input as model；
Whether each word is text after final correct in mark user-oriented dictionary library.If so, 1 is labeled as, if not being then labeled as 0；
According to each mark sample, using correlation machine learning algorithm, generate judge whether be text after candidate error correction model.
CN201810135187.4A 2018-02-09 2018-02-09 A kind of speech recognition text error correction method and device Pending CN108304385A (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201810135187.4A CN108304385A (en) 2018-02-09 2018-02-09 A kind of speech recognition text error correction method and device
PCT/CN2019/070817 WO2019153996A1 (en) 2018-02-09 2019-01-08 Text error correction method and apparatus for voice recognition

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810135187.4A CN108304385A (en) 2018-02-09 2018-02-09 A kind of speech recognition text error correction method and device

Publications (1)

Publication Number Publication Date
CN108304385A true CN108304385A (en) 2018-07-20

Family

ID=62864880

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810135187.4A Pending CN108304385A (en) 2018-02-09 2018-02-09 A kind of speech recognition text error correction method and device

Country Status (2)

Country Link
CN (1) CN108304385A (en)
WO (1) WO2019153996A1 (en)

Cited By (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109325227A (en) * 2018-09-14 2019-02-12 北京字节跳动网络技术有限公司 Method and apparatus for generating amendment sentence
CN109376362A (en) * 2018-11-30 2019-02-22 武汉斗鱼网络科技有限公司 A kind of the determination method and relevant device of corrected text
CN109614621A (en) * 2018-12-11 2019-04-12 中国移动通信集团江苏有限公司 A kind of method, device and equipment correcting text
CN109948144A (en) * 2019-01-29 2019-06-28 汕头大学 A method of the Teachers ' Talk Intelligent treatment based on classroom instruction situation
WO2019153996A1 (en) * 2018-02-09 2019-08-15 叶伟 Text error correction method and apparatus for voice recognition
CN110245606A (en) * 2019-06-13 2019-09-17 广东小天才科技有限公司 A kind of text recognition method, device, equipment and storage medium
CN110399607A (en) * 2019-06-04 2019-11-01 深思考人工智能机器人科技（北京）有限公司 A kind of conversational system text error correction system and method based on phonetic
CN110647987A (en) * 2019-08-22 2020-01-03 腾讯科技（深圳）有限公司 Method and device for processing data in application program, electronic equipment and storage medium
CN110765763A (en) * 2019-09-24 2020-02-07 金蝶软件（中国）有限公司 Error correction method and device for speech recognition text, computer equipment and storage medium
CN110782892A (en) * 2019-10-25 2020-02-11 四川长虹电器股份有限公司 Voice text error correction method
CN110782881A (en) * 2019-10-25 2020-02-11 四川长虹电器股份有限公司 Video entity error correction method after speech recognition and entity recognition
CN111326160A (en) * 2020-03-11 2020-06-23 南京奥拓电子科技有限公司 Speech recognition method, system and storage medium for correcting noise text
CN111339757A (en) * 2020-02-13 2020-06-26 上海凯岸信息科技有限公司 Error correction method for voice recognition result in collection scene
CN111401012A (en) * 2020-03-09 2020-07-10 北京声智科技有限公司 Text error correction method, electronic device and computer readable storage medium
CN111753531A (en) * 2020-06-28 2020-10-09 平安科技（深圳）有限公司 Text error correction method and device based on artificial intelligence, computer equipment and storage medium
CN111985234A (en) * 2020-09-08 2020-11-24 四川长虹电器股份有限公司 Voice text error correction method
CN112133295A (en) * 2020-11-09 2020-12-25 北京小米松果电子有限公司 Speech recognition method, apparatus and storage medium
CN112489655A (en) * 2020-11-18 2021-03-12 元梦人文智能国际有限公司 Method, system and storage medium for correcting error of speech recognition text in specific field
CN113012705A (en) * 2021-02-24 2021-06-22 海信视像科技股份有限公司 Error correction method and device for voice text
CN111368506B (en) * 2018-12-24 2023-04-28 阿里巴巴集团控股有限公司 Text processing method and device
CN116136957A (en) * 2023-04-18 2023-05-19 之江实验室 Text error correction method, device and medium based on intention consistency

Families Citing this family (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110399608B (en) * 2019-06-04 2023-04-25 深思考人工智能机器人科技（北京）有限公司 Text error correction system and method for dialogue system based on pinyin
CN110765772A (en) * 2019-10-12 2020-02-07 北京工商大学 Text neural network error correction model after Chinese speech recognition with pinyin as characteristic
CN111090341A (en) * 2019-12-24 2020-05-01 科大讯飞股份有限公司 Input method candidate result display method, related equipment and readable storage medium
CN111783433A (en) * 2019-12-26 2020-10-16 北京沃东天骏信息技术有限公司 Text retrieval error correction method and device
CN111241814B (en) * 2019-12-31 2023-04-28 中移（杭州）信息技术有限公司 Error correction method and device for voice recognition text, electronic equipment and storage medium
CN111209740B (en) * 2019-12-31 2023-08-15 中移（杭州）信息技术有限公司 Text model training method, text error correction method, electronic device and storage medium
CN111274785B (en) * 2020-01-21 2023-06-20 北京字节跳动网络技术有限公司 Text error correction method, device, equipment and medium
CN111369996B (en) * 2020-02-24 2023-08-18 网经科技（苏州）有限公司 Speech recognition text error correction method in specific field
CN113763961B (en) * 2020-06-02 2024-04-09 阿里巴巴集团控股有限公司 Text processing method and device
CN111666768A (en) * 2020-06-10 2020-09-15 京东方科技集团股份有限公司 Chinese named entity recognition method and device and electronic equipment
CN111897953B (en) * 2020-07-08 2022-12-30 西北工业大学 Network media platform comment text classification labeling data correction method
CN112257437B (en) * 2020-10-20 2024-02-13 中国科学技术大学 Speech recognition error correction method, device, electronic equipment and storage medium
CN112241445B (en) * 2020-10-26 2023-11-07 竹间智能科技（上海）有限公司 Labeling method and device, electronic equipment and storage medium
CN112528663B (en) * 2020-12-18 2024-02-20 中国南方电网有限责任公司 Text error correction method and system in power grid field scheduling scene
CN112685550B (en) * 2021-01-12 2023-08-04 腾讯科技（深圳）有限公司 Intelligent question-answering method, intelligent question-answering device, intelligent question-answering server and computer readable storage medium
CN112885335B (en) * 2021-01-22 2023-05-02 北京读我科技有限公司 Speech recognition method and related device
CN113823274B (en) * 2021-08-16 2023-10-27 华南理工大学 Voice keyword sample screening method based on detection error weighted editing distance
CN113674743A (en) * 2021-08-20 2021-11-19 云知声（上海）智能科技有限公司 ASR result replacement processing device and processing method used in natural language processing
CN114398463B (en) * 2021-12-30 2023-08-11 南京硅基智能科技有限公司 Voice tracking method and device, storage medium and electronic equipment

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150058007A1 (en) * 2013-08-26 2015-02-26 Samsung Electronics Co. Ltd. Method for modifying text data corresponding to voice data and electronic device for the same
CN105976818A (en) * 2016-04-26 2016-09-28 Tcl集团股份有限公司 Instruction identification processing method and apparatus thereof
CN107045496A (en) * 2017-04-19 2017-08-15 畅捷通信息技术股份有限公司 The error correction method and error correction device of text after speech recognition
CN107220235A (en) * 2017-05-23 2017-09-29 北京百度网讯科技有限公司 Speech recognition error correction method, device and storage medium based on artificial intelligence

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6859774B2 (en) * 2001-05-02 2005-02-22 International Business Machines Corporation Error corrective mechanisms for consensus decoding of speech
CN105374356B (en) * 2014-08-29 2019-07-30 株式会社理光 Audio recognition method, speech assessment method, speech recognition system and speech assessment system
CN108304385A (en) * 2018-02-09 2018-07-20 叶伟 A kind of speech recognition text error correction method and device

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150058007A1 (en) * 2013-08-26 2015-02-26 Samsung Electronics Co. Ltd. Method for modifying text data corresponding to voice data and electronic device for the same
CN105976818A (en) * 2016-04-26 2016-09-28 Tcl集团股份有限公司 Instruction identification processing method and apparatus thereof
CN107045496A (en) * 2017-04-19 2017-08-15 畅捷通信息技术股份有限公司 The error correction method and error correction device of text after speech recognition
CN107220235A (en) * 2017-05-23 2017-09-29 北京百度网讯科技有限公司 Speech recognition error correction method, device and storage medium based on artificial intelligence

Cited By (36)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019153996A1 (en) * 2018-02-09 2019-08-15 叶伟 Text error correction method and apparatus for voice recognition
CN109325227A (en) * 2018-09-14 2019-02-12 北京字节跳动网络技术有限公司 Method and apparatus for generating amendment sentence
US11531814B2 (en) 2018-09-14 2022-12-20 Beijing Bytedance Network Technology Co., Ltd. Method and device for generating modified statement
CN109376362A (en) * 2018-11-30 2019-02-22 武汉斗鱼网络科技有限公司 A kind of the determination method and relevant device of corrected text
CN109614621A (en) * 2018-12-11 2019-04-12 中国移动通信集团江苏有限公司 A kind of method, device and equipment correcting text
CN109614621B (en) * 2018-12-11 2023-09-19 中国移动通信集团江苏有限公司 Text correction method, device and equipment
CN111368506B (en) * 2018-12-24 2023-04-28 阿里巴巴集团控股有限公司 Text processing method and device
CN109948144A (en) * 2019-01-29 2019-06-28 汕头大学 A method of the Teachers ' Talk Intelligent treatment based on classroom instruction situation
CN109948144B (en) * 2019-01-29 2022-12-06 汕头大学 Teacher utterance intelligent processing method based on classroom teaching situation
CN110399607A (en) * 2019-06-04 2019-11-01 深思考人工智能机器人科技（北京）有限公司 A kind of conversational system text error correction system and method based on phonetic
CN110399607B (en) * 2019-06-04 2023-04-07 深思考人工智能机器人科技（北京）有限公司 Pinyin-based dialog system text error correction system and method
CN110245606A (en) * 2019-06-13 2019-09-17 广东小天才科技有限公司 A kind of text recognition method, device, equipment and storage medium
CN110245606B (en) * 2019-06-13 2021-07-20 广东小天才科技有限公司 Text recognition method, device, equipment and storage medium
CN110647987A (en) * 2019-08-22 2020-01-03 腾讯科技（深圳）有限公司 Method and device for processing data in application program, electronic equipment and storage medium
CN110765763A (en) * 2019-09-24 2020-02-07 金蝶软件（中国）有限公司 Error correction method and device for speech recognition text, computer equipment and storage medium
CN110765763B (en) * 2019-09-24 2023-12-12 金蝶软件（中国）有限公司 Error correction method and device for voice recognition text, computer equipment and storage medium
CN110782892A (en) * 2019-10-25 2020-02-11 四川长虹电器股份有限公司 Voice text error correction method
CN110782881A (en) * 2019-10-25 2020-02-11 四川长虹电器股份有限公司 Video entity error correction method after speech recognition and entity recognition
CN110782892B (en) * 2019-10-25 2022-03-25 四川长虹电器股份有限公司 Voice text error correction method
CN111339757A (en) * 2020-02-13 2020-06-26 上海凯岸信息科技有限公司 Error correction method for voice recognition result in collection scene
CN111401012A (en) * 2020-03-09 2020-07-10 北京声智科技有限公司 Text error correction method, electronic device and computer readable storage medium
CN111401012B (en) * 2020-03-09 2023-11-21 北京声智科技有限公司 Text error correction method, electronic device and computer readable storage medium
CN111326160A (en) * 2020-03-11 2020-06-23 南京奥拓电子科技有限公司 Speech recognition method, system and storage medium for correcting noise text
CN111753531B (en) * 2020-06-28 2024-03-12 平安科技（深圳）有限公司 Text error correction method, device, equipment and storage medium based on artificial intelligence
WO2021135444A1 (en) * 2020-06-28 2021-07-08 平安科技（深圳）有限公司 Text error correction method and apparatus based on artificial intelligence, computer device and storage medium
CN111753531A (en) * 2020-06-28 2020-10-09 平安科技（深圳）有限公司 Text error correction method and device based on artificial intelligence, computer equipment and storage medium
CN111985234B (en) * 2020-09-08 2022-02-01 四川长虹电器股份有限公司 Voice text error correction method
CN111985234A (en) * 2020-09-08 2020-11-24 四川长虹电器股份有限公司 Voice text error correction method
CN112133295A (en) * 2020-11-09 2020-12-25 北京小米松果电子有限公司 Speech recognition method, apparatus and storage medium
CN112133295B (en) * 2020-11-09 2024-02-13 北京小米松果电子有限公司 Speech recognition method, device and storage medium
CN112489655A (en) * 2020-11-18 2021-03-12 元梦人文智能国际有限公司 Method, system and storage medium for correcting error of speech recognition text in specific field
CN112489655B (en) * 2020-11-18 2024-04-19 上海元梦智能科技有限公司 Method, system and storage medium for correcting voice recognition text error in specific field
CN113012705B (en) * 2021-02-24 2022-12-09 海信视像科技股份有限公司 Error correction method and device for voice text
CN113012705A (en) * 2021-02-24 2021-06-22 海信视像科技股份有限公司 Error correction method and device for voice text
CN116136957A (en) * 2023-04-18 2023-05-19 之江实验室 Text error correction method, device and medium based on intention consistency
CN116136957B (en) * 2023-04-18 2023-07-07 之江实验室 Text error correction method, device and medium based on intention consistency

Also Published As

Publication number Publication date
WO2019153996A1 (en) 2019-08-15

Similar Documents

Publication Publication Date Title
CN108304385A (en) A kind of speech recognition text error correction method and device
CN110096570B (en) Intention identification method and device applied to intelligent customer service robot
US11848008B2 (en) Artificial intelligence-based wakeup word detection method and apparatus, device, and medium
US10614795B2 (en) Acoustic model generation method and device, and speech synthesis method
JP6198872B2 (en) Detection of speech syllable / vowel / phoneme boundaries using auditory attention cues
CN105869634B (en) It is a kind of based on field band feedback speech recognition after text error correction method and system
CN111680090B (en) Text structuring method, device and system and non-volatile storage medium
JP6189970B2 (en) Combination of auditory attention cue and phoneme posterior probability score for sound / vowel / syllable boundary detection
US10157619B2 (en) Method and device for searching according to speech based on artificial intelligence
CN107633851B (en) Discrete speech emotion recognition method, device and system based on emotion dimension prediction
CN106649742A (en) Database maintenance method and device
CN110853628A (en) Model training method and device, electronic equipment and storage medium
CN109034203A (en) Training, expression recommended method, device, equipment and the medium of expression recommended models
CN107122492A (en) Lyric generation method and device based on picture content
CN111180025A (en) Method and device for representing medical record text vector and inquiry system
CN107507613B (en) Scene-oriented Chinese instruction identification method, device, equipment and storage medium
Kowtha et al. Detecting emotion primitives from speech and their use in discerning categorical emotions
Narang et al. Abusive language detection using syntactic dependency graphs
CN107193941A (en) Story generation method and device based on picture content
CN112151021A (en) Language model training method, speech recognition device and electronic equipment
Huang et al. Detecting Intelligibility by Linear Dimensionality Reduction and Normalized Voice Quality Hierarchical Features.
CN114996466A (en) Method and system for establishing medical standard mapping model and using method
JP7207571B2 (en) LEARNING DATA GENERATION METHOD, LEARNING DATA GENERATION DEVICE, AND PROGRAM
CN114911922A (en) Emotion analysis method, emotion analysis device and storage medium
Vydana et al. Significance of emotionally significant regions of speech for emotive to neutral conversion

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
TA01 Transfer of patent application right

Effective date of registration: 20190402

Address after: Room 1287, 1/1, 8 Block 33 Guangshun Road, Changning District, Shanghai, 2003

Applicant after: Shanghai Century Network Technology Co., Ltd.

Address before: 200126 Yaohua Road, Pudong New Area, Shanghai, Room 204, room 560

Applicant before: Ye Wei

TA01 Transfer of patent application right
RJ01 Rejection of invention patent application after publication

Application publication date: 20180720

RJ01 Rejection of invention patent application after publication