CN110032740A - It customizes individual character semanteme and learns application method - Google Patents
It customizes individual character semanteme and learns application method Download PDFInfo
- Publication number
- CN110032740A CN110032740A CN201910320548.7A CN201910320548A CN110032740A CN 110032740 A CN110032740 A CN 110032740A CN 201910320548 A CN201910320548 A CN 201910320548A CN 110032740 A CN110032740 A CN 110032740A
- Authority
- CN
- China
- Prior art keywords
- information
- recognition result
- data library
- personalization database
- common denominator
- 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
- 238000000034 method Methods 0.000 title claims abstract description 21
- GNFTZDOKVXKIBK-UHFFFAOYSA-N 3-(2-methoxyethoxy)benzohydrazide Chemical compound COCCOC1=CC=CC(C(=O)NN)=C1 GNFTZDOKVXKIBK-UHFFFAOYSA-N 0.000 claims description 4
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000006243 chemical reaction Methods 0.000 abstract description 2
- 238000012790 confirmation Methods 0.000 description 5
- 206010011878 Deafness Diseases 0.000 description 4
- 238000004891 communication Methods 0.000 description 4
- 235000012054 meals Nutrition 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
It customizes individual character semanteme and learns application method, comprising the following steps: a) define personalization database;B) common denominator data library is defined;C) it inputs information and personalization database is called to be identified;D) operation is executed according to personalization database recognition result;E) it inputs information and common denominator data library is called to be identified;F) operation is executed according to common denominator data library recognition result;G) storage recognition result is chosen whether to personalization database, recognition result is such as stored, executes step a, if do not stored recognition result, executes step h;H) corresponding semantic information is exported.Compared with prior art, beneficial effects of the present invention: sound, image, movement etc. can be converted to corresponding semantic information, pass through customized semanteme, it can be by distinctive sound, image and action recognition at standard semantic information, the semantic conversion being able to achieve between different types of information, it can effectively improve discrimination and recognition speed, reduce error rate, realize personalized identification.
Description
Technical field
The present invention relates to semantics recognition field more particularly to personalized customization semantics recognition learning areas.
Background technique
With the continuous progress of science and technology, semantics recognition technology starts more and more to apply on various intelligent terminals, but
Be limited to the various factors such as processor performance, algorithm model, network bandwidth, current semantics recognition primarily directed to standard words and
A small number of dialects, it is clear to various dialects or asophia and the complex situations such as phonetic representation can not be passed through, it may appear that discrimination
Low, error rate is high, and recognition speed is slow, can not personalized identification the problems such as.
Summary of the invention
The present invention can carry out personalized customization individual character semanteme study application method in view of the above-mentioned problems, providing one kind,
Technical solution is as follows:
It customizes individual character semanteme and learns application method, comprising the following steps:
A) personalization database is defined;
B) common denominator data library is defined;
C) it inputs information and personalization database is called to be identified;
D) operation is executed according to personalization database recognition result;
D1) when correct identification, step h is executed;
D2) when part identifies or can not identify, step e is executed;
E) it inputs information and common denominator data library is called to be identified;
F) operation is executed according to common denominator data library recognition result:
When correct identification, step h is executed;
When part identifies, correct information is screened, executes step g;
When that can not identify, customized, execution step g is carried out to input information.
G) storage recognition result is chosen whether to personalization database:
Recognition result is stored, step a is executed;
Recognition result is not stored, step h is executed;
H) corresponding semantic information is exported.
Definition personalization database in step a is input self-defined information as information in personalization database.
Personalization database includes sound, writings and image corresponding informance.
Definition common denominator data library in step b is using existing database as common denominator data library.
Common denominator data library includes sound, writings and image corresponding informance.
The language (including dialect, unintelligible pronunciation, special sound or various movements) for identifying and learning user, generates a
Whether property database can prompt to store when identifying new voice or movement every time, not with frequency of use and the data of storage
Disconnected to increase, the scene domain covered also constantly expands, and discrimination is also higher.By taking Chinese as an example, when the data accumulation of storage is arrived
When the Chinese characters in common use of 2000-3000, most usage scenarios can be covered, if special screne needs, professional art can also be added
Language, popular word etc., then system can be converted to corresponding grapholect and voice automatically, further according to needing to translate into foreign language,
To meet the communication needs of the various language of people.
Compared with prior art, beneficial effects of the present invention: sound, image, movement etc. can be converted to corresponding language
Adopted information can be able to achieve difference by distinctive sound, image and action recognition at standard semantic information by customized semanteme
Semantic conversion between type information can effectively improve discrimination and recognition speed, reduce error rate, realize personalized knowledge
Not.
Detailed description of the invention
Fig. 1 is flow chart of the invention.
Specific embodiment
Embodiment 1
User first creates personalization database and common denominator data library, wherein the corresponding informance comprising sound, writings and image, individual character
Database needs user's self-defining, and existing database can be used in common denominator data library, when user needs to express semantic information
When, inputting information for the first time first can call personalization database to be identified, such as correct identification, then direct output information;Such as can not
Identification can call common denominator data library to be identified automatically, such as correct identification, then direct output information;If part identifies, user
It needs to confirm and identify whether accurately, and choose whether for recognition result to be stored in personalization database, semantic information is exported after confirmation;
It can not such as identify, user needs Manual definition's semantic information, and chooses whether for recognition result to be stored in personalization database, really
Semantic information is exported after recognizing;It is learning process that recognition result, which is stored in the process in personalization database, selectively storage identification
It as a result can be to avoid unwanted invalid information be generated in personalization database, when completion learning process and then secondary input are identical
When information, it can preferentially call the recognition result of personalization database storage to be identified, then export corresponding semantic information, then root
According to needing to be transcribed into multi-lingual and text.
Embodiment 2
User first creates personalization database and common denominator data library, wherein including voice and sign language and its corresponding text letter
Breath, personalization database need user's self-defining, and existing database can be used in common denominator data library, when user needs to input hand
When language movement is to express semanteme, when inputting the limb action information of oneself by image-input device for the first time, can preferentially it adjust
It is identified with personalization database, such as correct identification, then direct output information;Can not such as identify can call common denominator data library automatically
It is identified, such as correct identification, then direct output information;If part identifies, user, which needs to confirm, to be identified whether accurately, and is selected
It selects and whether recognition result is stored in personalization database, semantic information is exported after confirmation;It can not such as identify, user needs artificial
Semantic information is defined, and chooses whether for recognition result to be stored in personalization database, semantic information is exported after confirmation;Identification is tied
It is learning process that fruit, which is stored in the process in personalization database, and selectively store recognition result can be to avoid producing in personalization database
Raw unwanted invalid information, when inputting identical sign language information again, can preferentially call individual character after completing learning process
The recognition result of database purchase is identified, corresponding semantic information is then exported, multinational further according to needing to be transcribed into
Language and text.
Embodiment 3
When user needs to input oneself special sound or movement to express semanteme, sound and image-input device can be passed through
It is inputted, then its customized semanteme, customized semantic information is stored in personalization database, learning process is completed.When
When inputting special sound or movement again, it can preferentially call the recognition result of personalization database storage to be identified, then export
Corresponding semantic information, further according to needing to be transcribed into multi-lingual and text.
Embodiment 4
When occurring the text of mistake in identification process, manual confirmation is needed again to modify, for example, polyphone can be by Intelligent drainage
Sequence selects sequence number to determine, is defaulted as the first word of the pronunciation, is moved forward automatically according to frequency of use, with Sichuan Province China dialect and
For dialogue between texas,U.S dialect, this method can modify wrong word according to the word sequence number of display, such as say and " eat
When meal ", if it is shown that " 1 shame, 2 meal 34 scolding " as long as at this moment saying " 1 changes, and 4 change ", system can enter next error correction page
Face, display " 1-1 eat the 1-2 pond 1-3 1-4 hold 1-5 ruler 1-6 speed mother 4-1 4-2 4-3 4-4 code 4-5 scold ", then voice confirmation
1-1,4-3 can show selected text: " having had a meal " automatically, while can be automatically translated into the text and voice of English
To other side, it is also possible to Dezhou dialect (according to demands of individuals), other side on the contrary says Dezhou dialect also and can be converted any language
Text and deaf and dumb sign language, also may be implemented the Mixed design of voice, sign language and body action.
Embodiment 5
When deaf-mute and normal person link up, the sign language of deaf-mute is inputted by image input device, by calling data
Semanteme expressed by user is recognized in library, is then converted into the voice and text of standard, normal person is enable to understand its institute
The semanteme of expression, then the voice of normal person makes both sides again by the system converting sign language that can be identified at deaf-mute or text
It accessible can link up.
Embodiment 6
When needing to be linked up in particular circumstances, such as serious noise occasion forbids the occasion etc. of sound, the gesture of user or
Person's movement is inputted by image input device, semanteme expressed by user is identified by calling database, then by its turn
Become the text of standard, to realize written communication.
Embodiment 7
Various sound can be inputted in use and carry out storage record, and are identified as grapholect, such as the lyrics of song, flowing water
Sound, chirm, the sound of motor vehicle, tucket etc., making various sound all has corresponding semanteme.
Embodiment 8
It after special pattern is identified and defined in advance, stores into personalization database, when using special pattern, passes through figure
As input unit input, personalization database is called to be identified, and be converted into received pronunciation and text, to meet such as the disabled
The special communication requirements such as scholar.
Embodiment 9
Personalization database can be loaded into other application with modular form and be used, and such as be loaded into navigation application, be led with improving
Boat application meets the needs of people to the recognition efficiency of personal voice.
Embodiment 10
When user is under special state while moving (such as), the voice of sending generates variation, calls database can not positive common sense
It when other, will be operated according to first time voice input mode, user, which needs to confirm, to be identified whether accurately, by correct result
It is stored in personalization database, completes learning process.
Embodiment 11
The modes Mixed design information such as different language, sign language and figure may be implemented using the present invention, when people are multi-party long-range
In use, sender issues sign language information in video conference, recipient can voluntarily select intelligible text, voice and figure
Information receives, and realizes accessible communication.
Therefore, in all respects, the present embodiments are to be considered as illustrative and not restrictive, this
The range of invention is indicated by the appended claims rather than the foregoing description, it is intended that the equivalent requirements of the claims will be fallen in
All changes in meaning and scope are included within the present invention.It should not treat any reference in the claims as limitation institute
The claim being related to.
Claims (5)
1. customizing individual character semanteme learns application method, which comprises the following steps:
A) personalization database is defined;
B) common denominator data library is defined;
C) it inputs information and personalization database is called to be identified;
D) operation is executed according to personalization database recognition result;
D1) when correct identification, step h is executed;
D2) when part identifies or can not identify, step e is executed;
E) it inputs information and common denominator data library is called to be identified;
F) operation is executed according to common denominator data library recognition result:
When correct identification, step h is executed;
When part identifies, correct information is screened, executes step g;
When that can not identify, customized, execution step g is carried out to input information;
G) storage recognition result is chosen whether to personalization database:
Recognition result is stored, step a is executed;
Recognition result is not stored, step h is executed;
H) corresponding semantic information is exported.
2. the customization individual character semanteme learns application method according to claim 1, it is characterised in that determine in the step a
Adopted personalization database is input self-defined information as information in personalization database.
3. according to claim 1 or customization individual character semanteme described in 2 learns application method, it is characterised in that the personality data
Library includes sound, writings and image corresponding informance.
4. the customization individual character semanteme learns application method according to claim 1, it is characterised in that determine in the step b
Adopted common denominator data library is using existing database as common denominator data library.
5. according to claim 1 or customization individual character semanteme described in 4 learns application method, it is characterised in that the common denominator data
Library includes sound, writings and image corresponding informance.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910320548.7A CN110032740A (en) | 2019-04-20 | 2019-04-20 | It customizes individual character semanteme and learns application method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910320548.7A CN110032740A (en) | 2019-04-20 | 2019-04-20 | It customizes individual character semanteme and learns application method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110032740A true CN110032740A (en) | 2019-07-19 |
Family
ID=67239392
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910320548.7A Pending CN110032740A (en) | 2019-04-20 | 2019-04-20 | It customizes individual character semanteme and learns application method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110032740A (en) |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101527092A (en) * | 2009-04-08 | 2009-09-09 | 西安理工大学 | Computer assisted hand language communication method under special session context |
CN102831195A (en) * | 2012-08-03 | 2012-12-19 | 河南省佰腾电子科技有限公司 | Individualized voice collection and semantics determination system and method |
JP2015026057A (en) * | 2013-07-29 | 2015-02-05 | 韓國電子通信研究院Electronics and Telecommunications Research Institute | Interactive character based foreign language learning device and method |
CN106446836A (en) * | 2016-09-28 | 2017-02-22 | 戚明海 | Sign language recognition and interpretation device |
CN106649278A (en) * | 2016-12-30 | 2017-05-10 | 三星电子(中国)研发中心 | Method and system for extending spoken language dialogue system corpora |
CN108268835A (en) * | 2017-12-28 | 2018-07-10 | 努比亚技术有限公司 | sign language interpretation method, mobile terminal and computer readable storage medium |
CN108427910A (en) * | 2018-01-30 | 2018-08-21 | 浙江凡聚科技有限公司 | Deep-neural-network AR sign language interpreters learning method, client and server |
CN109215638A (en) * | 2018-10-19 | 2019-01-15 | 珠海格力电器股份有限公司 | A kind of phonetic study method, apparatus, speech ciphering equipment and storage medium |
-
2019
- 2019-04-20 CN CN201910320548.7A patent/CN110032740A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101527092A (en) * | 2009-04-08 | 2009-09-09 | 西安理工大学 | Computer assisted hand language communication method under special session context |
CN102831195A (en) * | 2012-08-03 | 2012-12-19 | 河南省佰腾电子科技有限公司 | Individualized voice collection and semantics determination system and method |
JP2015026057A (en) * | 2013-07-29 | 2015-02-05 | 韓國電子通信研究院Electronics and Telecommunications Research Institute | Interactive character based foreign language learning device and method |
CN106446836A (en) * | 2016-09-28 | 2017-02-22 | 戚明海 | Sign language recognition and interpretation device |
CN106649278A (en) * | 2016-12-30 | 2017-05-10 | 三星电子(中国)研发中心 | Method and system for extending spoken language dialogue system corpora |
CN108268835A (en) * | 2017-12-28 | 2018-07-10 | 努比亚技术有限公司 | sign language interpretation method, mobile terminal and computer readable storage medium |
CN108427910A (en) * | 2018-01-30 | 2018-08-21 | 浙江凡聚科技有限公司 | Deep-neural-network AR sign language interpreters learning method, client and server |
CN109215638A (en) * | 2018-10-19 | 2019-01-15 | 珠海格力电器股份有限公司 | A kind of phonetic study method, apparatus, speech ciphering equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111477216B (en) | Training method and system for voice and meaning understanding model of conversation robot | |
Merdivan et al. | Dialogue systems for intelligent human computer interactions | |
CN111223498A (en) | Intelligent emotion recognition method and device and computer readable storage medium | |
CN112352275A (en) | Neural text-to-speech synthesis with multi-level textual information | |
CN111339771B (en) | Text prosody prediction method based on multitasking multi-level model | |
CN112131359A (en) | Intention identification method based on graphical arrangement intelligent strategy and electronic equipment | |
CN116863038A (en) | Method for generating digital human voice and facial animation by text | |
CN112599113B (en) | Dialect voice synthesis method, device, electronic equipment and readable storage medium | |
CN114911932A (en) | Heterogeneous graph structure multi-conversation person emotion analysis method based on theme semantic enhancement | |
CN113628610A (en) | Voice synthesis method and device and electronic equipment | |
CN116303966A (en) | Dialogue behavior recognition system based on prompt learning | |
CN109933773A (en) | A kind of multiple semantic sentence analysis system and method | |
GB2376554A (en) | Artificial language generation and evaluation | |
CN112257432A (en) | Self-adaptive intention identification method and device and electronic equipment | |
KR20220070826A (en) | Utterance manipulation apparatus for retrieval-based response selection and method thereof | |
CN112489634A (en) | Language acoustic model training method and device, electronic equipment and computer medium | |
CN117012177A (en) | Speech synthesis method, electronic device, and storage medium | |
CN110032740A (en) | It customizes individual character semanteme and learns application method | |
CN112242134A (en) | Speech synthesis method and device | |
CN115374784A (en) | Chinese named entity recognition method based on multi-mode information selective fusion | |
CN112150103B (en) | Schedule setting method, schedule setting device and storage medium | |
CN115238048A (en) | Quick interaction method for joint chart identification and slot filling | |
CN114974218A (en) | Voice conversion model training method and device and voice conversion method and device | |
Fujimoto et al. | Semi-supervised learning based on hierarchical generative models for end-to-end speech synthesis | |
O'Brien | Knowledge-based systems in speech recognition: a survey |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190719 |