CN101282379A - Method for inputting special words through voice for mobile phone - Google Patents
Method for inputting special words through voice for mobile phone Download PDFInfo
- Publication number
- CN101282379A CN101282379A CNA2008101057073A CN200810105707A CN101282379A CN 101282379 A CN101282379 A CN 101282379A CN A2008101057073 A CNA2008101057073 A CN A2008101057073A CN 200810105707 A CN200810105707 A CN 200810105707A CN 101282379 A CN101282379 A CN 101282379A
- Authority
- CN
- China
- Prior art keywords
- mobile phone
- specific word
- input
- voice
- user
- 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
Images
Landscapes
- Telephonic Communication Services (AREA)
- Telephone Function (AREA)
Abstract
The invention provides a voice input method for cell phones, which is characterized in that: first, the voice input is based on words rather than sentences. Second, finite carefully-selected common words group are preset in the cell phone as input limit, only these pre-selected words are input by voice, the cell phone recognition system can correctly recognize user voice and transform the voice to corresponding text. Third, the well-chosen words group is preset before the cell phone leaves factory, which is not needed for users to create, but users can add in their extra common words. The voice input method for cell phones of the invention facilitates information input for users and improves the whole input speed.
Description
Technical field
Disclosed by the invention is a kind of method that is used for the phonetic entry of mobile phone, relates in particular to a kind ofly based on vocabulary input, and is a kind of method that limits the input lexical scoping.
Background technology
Mobile phone has more and more become product indispensable in people's life.It makes that people's communication is more smooth and easy.
For the input method of mobile phone, modal two kinds is inputting method and hand-writing input method at present.These two kinds of input methods can satisfy most of crowd's input demand.But, certain limitation is also arranged respectively.For example, the spelling input method that inputting method mainly comprises, for the people who is unfamiliar with phonetic, the non-type people that perhaps pronounces has very big difficulty.Difficulty and human brain and the inharmonic obstacle of pronunciation that the stroke method also exists some strokes to split be not so be used widely.And, just inputting method can't have been used for some mobile phones that does not have keyboard or have only several simple buttons.As long as the advantage of handwriting input is to write, just can imports, but must be able to write, because the difference of educational level also exists a lot of words can not write.In addition, for the mobile phone that does not have touch-screen, hand-written method is just useless.
Seen top two kinds of methods, people tend to propose whether can accomplish the user as long as say the language that oneself will import facing to mobile phone, and mobile phone converts it into text by recognizer.As long as mobile phone has Mike, just there is not other to limit like this.And well-done words, also can similarly discern dialect.But, this idea, is it much that we need look at earlier that feasibility has?
Same idea, the simplest relatively is exactly present developing faster computer, because the speed of service of computer is higher than mobile phone far away, the main cause mobile phone is a handheld device, need make an issue of on power consumption and property easy to carry.So the thing that can't finish is difficult on the low relatively platform of this disposal ability of mobile phone and realizes on computers.Remember at the beginning of 2008 that Microsoft had once mentioned the next generation of computer development, can carry out phonetic entry.But that is of future generation, needs algorithm and chip disposal ability significantly to improve synchronously.So the meaning between the lines is exactly present technology, can't realize real phonetic entry.So can not on the mobile phone yet.
The research work of Chinese speech input launches already, and very big progress has been arranged, although speaking and there is a big difference naturally from the identification people.But present technology can carry out feature extraction to the sound bite of the clauses and subclauses that limit and go to compare in limited entry set selecting the highest result of probability.Present technology can be done limited entry set increasing, and yes under the prerequisite that guarantees corresponding speed.
In addition, by the vocabulary of the normal input of people is studied, more professional research institution is arranged certainly, general situation is that the vocabulary that present people use always has about 3,000, these 3,000 speech can cover 86% of general language material, 5,000 everyday words can cover 91%, eight thousand and then can cover 95%.This is very soul-stirring statistics.We can pass through this, reduce the difficulty of speech recognition, reduce the quality of identification again not too bigly.
Based on above-mentioned consideration, expected in the speech recognition technology input method of the limited range of unspecified person.This is a theory, and just as people will realize speech recognition, the someone thought for a long time, but how to realize that the no one does basically.The present invention from this disclosed theoretical thinking, has selected suitable simplification just, reaches on the not strong equipment of the such disposal ability of mobile phone, finishes the input of the most of everyday expressions of people.This technology, abroad, major company especially, so the problem of possible worry about discrimination is inreal research and practical to mobile phone.This also is a main limitation of the present invention, promptly has certain discrimination problem.The level that can reach can satisfy most of human needs substantially about 90% at present, in the time of especially can't using keyboard and handwriting input.In addition, can not import all speech, also be the limitation of this method.But in conjunction with other input method of mobile phone, this pronunciation inputting method can play extraordinary effect.And on the speed than keyboard and hand-written fast, differ certainly and compared with the keyboard fan surely.
Summary of the invention
For addressing this problem, the present invention proposes a kind of be used for mobile phone can carry out the method for phonetic entry to specific word, thereby introduced a kind of very convenient effective input method.
In mobile phone, need to carry out the place of text input, the user starts speech recognition system by a certain dedicated array of keys, mobile phone prompting user reads the text that will import, after the user speech input is finished, system extracts feature to the voice of input, then with mobile phone in the feature of existing specific word compare one by one, find the most close clauses and subclauses, and with its corresponding vocabulary text feedback in the place that mobile phone carries out text input, feed back to the user.If differ greatly, can the feedback search failure.
In the above process, carry out the place of text input, be meant to comprise note in the mobile phone, multimedia message, Email, notepad, the place of carrying out the text input of functions such as schedule.
The startup of above-described speech recognition system is the user in the place of text input, by carrying out after dedicated array of keys.Dedicated array of keys can be any one button on the mobile phone, includes but not limited to directionkeys, acknowledgement key, the button of side, numerical key, function key, left and right sides soft key etc.
Above-described prompting user reads the text that will import, and the mode of its prompting can be to point out with voice suggestion or with videotex.
Above-described system extracts feature to the voice of input, and this is a wide in range relatively notion, and promptly the feature extracting methods that may use in fields such as speech recognition and artificial intelligence at present is abstracted into feature with user's voice.
Existing specific word in the above-described mobile phone is meant that mobile phone producer just is preset at the speech that producer presets that is different from that specific word in the mobile phone and user import before mobile phone dispatches from the factory after buying mobile phone.And described specific word, the number restriction is arranged, following two aspects have mainly been considered in the restriction of number: at first, based on the processing speed of present handset processes chip and the optimization degree of recognizer; Secondly, based in the linguistics to the statistics of common words.The ceiling restriction of the number that the present invention requires is in 2000 to 6000 scopes.Be to be limited to 2000 to 6000 on the number of specific word.The number of the specific word in this method does not have lower limit.And number is few more, and discrimination is high more.In addition, described specific word has the restriction of literal number, and the present invention requires for each specific word, and the number of its Chinese text is that maximum is no more than 4.
Above-describedly find the most close clauses and subclauses, be meant the most close one of phonetic feature value, if but the difference of the most close characteristic value exceeds certain limit, and also can return and search failure, the prompting user can not find.Certainly, if close several all differences are little, can also provide the more candidate item of user selective.
Above-described feedback both was meant and can comes display result with display screen in the place that mobile phone carries out the text input, also can the vocabulary text reading of discerning be come out by the mode of massage voice reading.
Description of drawings
Fig. 1 is an overall flow schematic diagram of the present invention;
Embodiment
Further specify embodiments of the present invention below in conjunction with example and accompanying drawing.
Overall flow schematic diagram of the present invention, as shown in Figure 1.Concrete implementation step is as follows:
Step 1, the user enters the place that needs to carry out the text input in the mobile phone, note for example, multimedia message, Email, notepad, the place of carrying out the text input of functions such as schedule.
Step 2, the user starts speech recognition system by a certain dedicated array of keys, and dedicated array of keys can be any one button on the mobile phone, includes but not limited to directionkeys, acknowledgement key, the button of side, numerical key, function key, left and right sides soft key etc.
Step 3, mobile phone prompting user reads the text that will import.The mode of prompting can be to point out with voice suggestion or with videotex.
Step 4, after the user speech input was finished, system extracted feature to the voice of input, and this is a wide in range relatively notion, and promptly the feature extracting methods that may use in fields such as speech recognition and artificial intelligence at present is abstracted into feature with user's voice.
Step 5 is compared one by one with the feature of existing specific word in the mobile phone.Existing specific word in the mobile phone is meant that mobile phone producer just is preset at the speech that producer presets that is different from that specific word in the mobile phone and user import before mobile phone dispatches from the factory after buying mobile phone.And described specific word, the number restriction is arranged, following two aspects have mainly been considered in the restriction of number: at first, based on the processing speed of present handset processes chip and the optimization degree of recognizer; Secondly, based in the linguistics to the statistics of common words.The ceiling restriction of the number that the present invention requires is in 2000 to 6000 scopes.Be to be limited to 2000 to 6000 on the number of specific word.The number of the specific word in this method does not have lower limit.And number is few more, and discrimination is high more.In addition, described specific word has the restriction of literal number, and the present invention requires for each specific word, and the number of its Chinese text is that maximum is no more than 4.
Step 6 finds the most close clauses and subclauses, i.e. the most close one of phonetic feature value, and with its corresponding vocabulary text feedback in the place that mobile phone carries out the text input, feed back to the user.If but the difference of the most close characteristic value exceeds certain limit, also can return and search failure, the prompting user can not find.Feedback can be come display result with display screen, also can the vocabulary text reading of discerning be come out by the mode of massage voice reading.Certainly, if close several all differences are little, can also provide the more candidate item of user selective.
Claims (10)
1, a kind of method that can carry out phonetic entry to specific word that is used for mobile phone is characterized in that:
In mobile phone, need to carry out the place of text input, the user starts speech recognition system by a certain dedicated array of keys, mobile phone prompting user reads the text that will import, after the user speech input is finished, system extracts feature to the voice of input, then with mobile phone in the feature of existing specific word compare one by one, find the most close clauses and subclauses, and with its corresponding vocabulary text feedback in the place that mobile phone carries out text input, feed back to the user.If differ greatly, can the feedback search failure.
2, method according to claim 1 is characterized in that:
Described needs carry out the place of text input, are meant to comprise note in the mobile phone multimedia message, Email, notepad, the place of carrying out the text input of functions such as schedule.
3, method according to claim 1 is characterized in that:
The startup of described speech recognition system is the user in the place of text input, by carrying out after dedicated array of keys.Dedicated array of keys can be any one button on the mobile phone, includes but not limited to directionkeys, acknowledgement key, the button of side, numerical key, function key, left and right sides soft key etc.
4, method according to claim 1 is characterized in that:
Described prompting user reads the text that will import, and the mode of its prompting can be to point out with voice suggestion or with videotex.
5, method according to claim 1 is characterized in that:
Described system extracts feature to the voice of input, and this is a wide in range relatively notion, and promptly the feature extracting methods that may use in fields such as speech recognition and artificial intelligence at present is abstracted into feature with user's voice.
6, method according to claim 1 is characterized in that:
Existing specific word in the described mobile phone is meant that mobile phone producer just is preset at the speech that producer presets that is different from that specific word in the mobile phone and user import before mobile phone dispatches from the factory after buying mobile phone.
7, according to claim 1 and 6 described specific word, it is characterized in that:
Described specific word has the number restriction, and the number that the present invention requires is limited in 2000 to 6000 scopes.Be to be limited to 2000 to 6000 on the number of specific word.The number of the specific word in this method does not have lower limit.And number is few more, and discrimination is high more.
8, according to claim 1 and 6 described specific word, it is characterized in that:
Described specific word has the restriction of literal number.The present invention requires for each specific word, and the number of its Chinese text is that maximum is no more than 4.
9, method according to claim 1 is characterized in that:
Describedly find the most close clauses and subclauses, be meant the most close one of phonetic feature value, if but the difference of the most close characteristic value exceeds certain limit, and also can return and search failure, the prompting user can not find.Certainly, if close several all differences are little, can also provide the more candidate item of user selective.
10, method according to claim 1 is characterized in that:
Described feedback both was meant and can comes display result with display screen in the place that mobile phone carries out the text input, also can the vocabulary text reading of discerning be come out by the mode of massage voice reading.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2008101057073A CN101282379A (en) | 2008-04-30 | 2008-04-30 | Method for inputting special words through voice for mobile phone |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2008101057073A CN101282379A (en) | 2008-04-30 | 2008-04-30 | Method for inputting special words through voice for mobile phone |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101282379A true CN101282379A (en) | 2008-10-08 |
Family
ID=40014651
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2008101057073A Pending CN101282379A (en) | 2008-04-30 | 2008-04-30 | Method for inputting special words through voice for mobile phone |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101282379A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102902506A (en) * | 2012-09-18 | 2013-01-30 | 百度在线网络技术（北京）有限公司 | Voice input method and device of mobile terminal |
| CN105161111A (en) * | 2015-08-25 | 2015-12-16 | 百度在线网络技术（北京）有限公司 | Bluetooth connection-based voice recognition method and device |
-
2008
- 2008-04-30 CN CNA2008101057073A patent/CN101282379A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102902506A (en) * | 2012-09-18 | 2013-01-30 | 百度在线网络技术（北京）有限公司 | Voice input method and device of mobile terminal |
| CN102902506B (en) * | 2012-09-18 | 2016-01-06 | 百度在线网络技术（北京）有限公司 | Pronunciation inputting method in a kind of mobile terminal and device |
| CN105161111A (en) * | 2015-08-25 | 2015-12-16 | 百度在线网络技术（北京）有限公司 | Bluetooth connection-based voice recognition method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020238045A1 (en) | Intelligent speech recognition method and apparatus, and computer-readable storage medium | |
| WO2005101235A1 (en) | Dialogue support device | |
| CN100592385C (en) | Method and system for performing speech recognition on multi-language name | |
| Yağanoğlu | Real time wearable speech recognition system for deaf persons | |
| CN104202455A (en) | Intelligent voice dialing method and intelligent voice dialing device | |
| CN101105894A (en) | Multifunctional language learning machine | |
| Panda | Automated speech recognition system in advancement of human-computer interaction | |
| CN102063282B (en) | Chinese speech input system and method | |
| CN112669842A (en) | Man-machine conversation control method, device, computer equipment and storage medium | |
| CN110675866A (en) | Method, apparatus and computer-readable recording medium for improving at least one semantic unit set | |
| CN112489634A (en) | Language acoustic model training method and device, electronic equipment and computer medium | |
| Fellbaum et al. | Principles of electronic speech processing with applications for people with disabilities | |
| CN102970618A (en) | Video on demand method based on syllable identification | |
| JP2009515227A (en) | Text input system and method based on speech recognition | |
| CN101282379A (en) | Method for inputting special words through voice for mobile phone | |
| CN102262524A (en) | Method for recognizing and inputting sound based on wireless input equipment and device for implementing method | |
| CN1963753A (en) | Multidimensional Chinese understanding system and method applied for Chinese study | |
| US20190304454A1 (en) | Information providing device, information providing method, and recording medium | |
| CN107251137A (en) | Improve method, device and the computer readable recording medium storing program for performing of the set of at least one semantic primitive using voice | |
| Venkatagiri | Speech recognition technology applications in communication disorders | |
| CN1275174C (en) | Chinese language input method possessing speech sound identification auxiliary function and its system | |
| JP2002268680A (en) | Hybrid oriental character recognition technology using key pad and voice in adverse environment | |
| CN1965349A (en) | Multimodal disambiguation of speech recognition | |
| JP2004170466A (en) | Voice recognition method and electronic device | |
| Hernandez-Rebollar | Gesture-driven American sign language phraselator |
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 |
Application publication date: 20081008 |