CN105117034A - Method for inputting Chinese speeches, positioning statements and correcting errors - Google Patents
Method for inputting Chinese speeches, positioning statements and correcting errors Download PDFInfo
- Publication number
- CN105117034A CN105117034A CN201510557031.1A CN201510557031A CN105117034A CN 105117034 A CN105117034 A CN 105117034A CN 201510557031 A CN201510557031 A CN 201510557031A CN 105117034 A CN105117034 A CN 105117034A
- Authority
- CN
- China
- Prior art keywords
- chinese character
- sequence
- chinese
- statement
- syllable
- 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
Abstract
The invention discloses a method for inputting Chinese speeches, positioning statements and correcting errors. The method for inputting Chinese speeches, positioning statements and correcting errors is characterized by comprising the steps of converting human language into keyboard language (namely, converting speech input into Pinyin input), synchronously recognizing the statement lines of the Chinese characters, cached on an input method interface, and carrying out digital sequence positioning and error correction on the Chinese characters and Pinyin syllables in the lines, wherein the speech input flow comprises the processes of speech filtering, speech correction, language conversion, Chinese character recognition, sequence positioning, Chinese character error correction and the like.
Description
Technical field:
The present invention is applied to the platform of desktop computer, notebook computer, mobile phone, panel computer and other phonetic entries.
Background technology:
Along with the develop rapidly of phonetic entry technology and the popularization and application on mobile phone, people, for typewriting technology, more and more favor and pursue the boundary of [A gentleman uses his tongue but not his firsts].Current phonetic entry law technology is voice entry system substantially after the voice obtaining people, directly carries out decoding and obtains word, then feed back to user to voice signal.The input pattern of this tone decoding, its result of voice analysis is limited by speech speed, the dead time, thus causes identifying rate variance, and accuracy rate is not high.And the word that phonetic entry obtains after resolving directly enters in document, and when there is Text region mistake, amendment is wasted time and energy.
User's accent is corrected as mandarin standard pronunciation by adopting by the present invention, be converted to the keyboard Chinese phonetic alphabet again, its result of voice analysis is equivalent to Chinese phonetic alphabet typewriting, it is not by speech speed and the impact of dead time, add and utilize system phonetic words storehouse, corpus and cloud input technology, improve the accuracy rate of speech recognition greatly; Meanwhile, the present invention adopts interface of input method, and the word of phonetic entry conversion is not directly go up screen document but be buffered in the statement line of interface of input method, and adopts numeral mark sequence location to statement Chinese character, facilitates the amendment of speech recognition errors greatly.
Summary of the invention:
A kind of Chinese speech input method and statement are located and error correction method, it is characterized in that: human language is synchronously changed into keyboard language (being synchronously converted to Pinyin Input by phonetic entry) and synchronously identify Chinese character and be buffered in interface of input method statement line, and by digital sequence, position is demarcated to statement line Chinese character and pinyin syllable and corrects a mistake, phonetic entry flow process comprises voice filtering, phoniatrics, language conversion, Chinese Character Recognition, sequence demarcate the process such as position, Chinese character error correction, wherein:
The function of voice filtering procedure is: by the voice transmitted by microphone (comprising ambient noise and noise), by arranging the threshold value of the various technical parameter such as amplitude, frequency, wavelength, the sound, tone, tone color of sound, filtering various noise and noise and screening out the voice (namely implementing the accent of input in Chinese actor) that current line behaves.
The function of phoniatrics process is: the standard mandarin voice by non-standard mandarin pronunciation (i.e. the accent of the involved party) synchronous correction of involved party being system intialization.
The function of language conversion process is: human language is synchronously converted to keyboard language, and the standard mandarin voice flow by people is synchronously converted to keyboard Chinese Pinyin syllables sequence.
The function of Chinese Character Recognition process is: the synchronous contrast Chinese phonetic alphabet identifies Chinese character and outputs to interface of input method statement line.
The function of sequence mark position fixing process is: immediately below the Chinese character and syllable of statement line, demarcate position mark sequence by digital sequence.
The function of Chinese character error correction procedure is: application sequence demarcates bit function, corrects the Chinese character of identification error in phonetic entry.
Use phonitic entry method, first want opening voice interface of input method, interface is made up of statement line and the candidate Chinese character space of a whole page.
Statement line is rectangular frame, point line of text and phonetic capable, line of text buffer memory statement Chinese character, phonetic row cache statement pinyin syllable sequence, the Chinese character that phonetic entry identification converts is not directly go up screen but first buffer memory is presented in the statement line of interface of input method, statement line can show 20 Chinese characters, and when while statement is more than 20 Chinese characters, Chinese character above can head into screen automatically.Press [space] or [carriage return] key, can defeated for statement line word upper screen.
In statement line buffer memory display Chinese character, system can with numeral immediately below Chinese character location mark sequence, front 10 Chinese characters with numeral 1., 2., 3., 4., 5., 6., 7., 8., 9.,
sortord location mark sequence, rear 10 Chinese characters are with numeral
sortord location mark sequence; (this programme, in concrete software input method, front 10 digital sequence marks also can identify with green 1234567890, and rear 10 digital sequence marks also can adopt red 1234567890 to identify).
The Chinese phonetic alphabet in statement line, splits naturally with syllable separator [], and a syllable corresponds to a Chinese character, and a Chinese character corresponds to a syllable, and system also locates mark sequence with identical sequence mark to pinyin syllable while to Chinese character location mark sequence.
The candidate Chinese character space of a whole page is separate-grid type square box, is positioned at above statement line, the prepare word of display current speech.
, can there are two kinds of mistakes in phonetic entry: one is that spoken voice rectification makes a mistake for during mandarin standard pronunciation, thus causes the Chinese phonetic alphabet mistake of conversion, the Chinese character mistake of identification unavoidably; One is that mandarin standard pronunciation phonetically similar word mistake occurs when converting the Chinese phonetic alphabet to.When phonetic entry generation Chinese Character Recognition mistake, following method can be adopted to revise:
Method one, statement displacement amendment: by the sequence label of word before erroneous words, screen on all correct word allowing before erroneous words, mark sequence reorientated in remaining Chinese character, and meanwhile, cursor highlights and is positioned in erroneous words, choose correct candidate word and substitute erroneous words, or re-enter correct voice.
1. sequence is designated as, 2., 3., 4., 5., 6., 7., 8., 9.,
chinese character, screen on the numerical key on direct keypad; Sequence is designated as
chinese character, by [0+ numeral] upper screen.
Method two, cursor location amendment: mobile cursor, navigates to the amendment of erroneous words place, and when cursor location Chinese character, Chinese character can highlight display.
Method three, sequence demarcate position amendment: 1. sequence is designated as, 2., 3., 4., 5., 6., 7., 8., 9.,
chinese character, by [Ctrl+ numeral], a key quick position is to erroneous words place; Sequence is designated as
chinese character, by [Alt+ numeral], a key quick position is to erroneous words place.
Adopt the amendment of cursor location and sequence to demarcate position amendment, cursor is in location and also locate while highlighting Chinese character and highlight pinyin syllable corresponding to Chinese character.
This programme be used for Chinese character and syllable location mark sequence sequence mark, be not limited to 1., 2., 3., 4., 5., 6., 7., 8., 9.,
sequence, can adopt other digital forms, letter or symbol combination mark sequence, and 1,2,3,4,5,6,7,8,9,0 also can be adopted to assign to mark sequence with the chromatic zones that front and back group is different simultaneously; Sequence target position, is not limited to immediately below Chinese character and syllable, also can directly over, or subscript or subscript.
Accompanying drawing explanation
Fig. 1: phonetic entry process flow diagram.
Fig. 2: phonetic entry process interface of input method schematic diagram.
Fig. 3: statement is revised---statement is shifted.
Fig. 4: statement is revised---cursor location or sequence demarcate position.
Embodiment:
Use phonetic entry, necessary first opening voice input method, display input method interface.Press [space] or [carriage return] key, can defeated for statement line word upper screen document.
When user says [Chinese character of phonetic entry is not directly upper the screen but buffer memory statement line] the words facing to microphone, the accent transmitted by microphone is first synchronously converted to standard mandarin sound by technology such as filtration, rectifications by phonitic entry method system, and then be synchronously converted to the keyboard Chinese phonetic alphabet, utilize the technology such as phonetic character library, corpus, cloud input to identify concrete Chinese character output in the statement line of interface of input method, and input method system can to statement Chinese character and syllable location mark sequence (see [Fig. 2]) simultaneously.
The words was above input [upper screen] word originally, may be because do not have [upper screen] this word in system dictionary, or there is [upper level ground] this ground noun, input method has been identified as [upper level ground] [upper screen], there occurs the identification of mistake, at this moment, can revise in the following way:
A kind of mode is statement displacement amendment: the location sequence mark of wrongly written character [level ground] is
sequence mark
all Chinese characters are above all correct, therefore, by [0+2],
correct above word [Chinese character of phonetic entry be not directly on] first sends into screen document, remaining Chinese character [level ground but buffer memory statement line] is allowed to reorientate mark sequence (after displacement, first Chinese character and syllable can highlight display), then select [screen] to replace [level ground] (see [Fig. 3]).
A kind of mode is the amendment of statement location: mobile cursor is to sequence mark
place, or by [Alt+3] quick position to sequence mark
place's (Chinese character behind location and syllable can highlight display), chooses [screen] and replaces [level ground] word (see [Fig. 4]).
After the amendment of mistake word, press [space] or [carriage return] key, statement line word is just input in screen document and has gone.
Claims (3)
1. a Chinese speech input method and statement are located and error correction method, it is characterized in that: human language is synchronously changed into keyboard language (being synchronously converted to Pinyin Input by phonetic entry) and synchronously identify Chinese character and be buffered in interface of input method statement line, and by digital sequence, position is demarcated to statement line Chinese character and pinyin syllable and corrects a mistake, input flow process comprises voice filtering, phoniatrics, language conversion, Chinese Character Recognition, sequence demarcate the process such as position, Chinese character error correction, wherein:
The function of voice filtering procedure is: by the voice transmitted by microphone, by arranging the threshold value of the various technical parameter such as amplitude, frequency, wavelength, the sound, tone, tone color of sound, filtering various noise and noise and screening out the voice that current line behaves;
The function of phoniatrics process is: the standard mandarin voice by the non-standard mandarin pronunciation synchronous correction of involved party being system intialization;
The function of language conversion process is: human language is synchronously converted to keyboard language, and the standard mandarin voice flow by people is synchronously converted to keyboard Chinese Pinyin syllables sequence;
The function of Chinese Character Recognition process is: the synchronous contrast Chinese phonetic alphabet identifies Chinese character and outputs in interface of input method statement line;
The function of sequence mark position fixing process is: immediately below the Chinese character and syllable of statement line, demarcate position mark sequence by digital sequence;
The function of Chinese character error correction procedure is: application sequence demarcates bit function, corrects the Chinese character of identification error in phonetic entry;
Use Chinese speech input method, first want opening voice interface of input method, interface of input method is made up of statement line and the candidate Chinese character space of a whole page;
Statement line is rectangular frame, point line of text and phonetic capable, line of text buffer memory statement Chinese character, phonetic row cache statement pinyin syllable sequence, the Chinese character that phonetic entry identification converts is not directly go up screen but first buffer memory is presented in the statement line of interface of input method, and statement line can show 20 Chinese characters, when while statement is more than 20 Chinese characters, Chinese character above can head into screen automatically, presses [space] or [carriage return] key, can defeated for statement line word upper screen;
In statement line buffer memory display Chinese character, system can with numeral immediately below Chinese character location mark sequence, front 10 Chinese characters with numeral 1., 2., 3., 4., 5., 6., 7., 8., 9.,
sortord location mark sequence, rear 10 Chinese characters are with numeral
sortord location mark sequence;
The Chinese phonetic alphabet in statement line, splits naturally with syllable separator, and a syllable corresponds to a Chinese character, and a Chinese character corresponds to a syllable, and system also locates mark sequence with identical sequence mark to pinyin syllable while to Chinese character location mark sequence;
The candidate Chinese character space of a whole page is separate-grid type square box, is positioned at above statement line, the prepare word of display current speech;
, can there are two kinds of mistakes in phonetic entry: one is that spoken voice rectification makes a mistake for during mandarin standard pronunciation, thus causes the Chinese phonetic alphabet mistake of conversion, the Chinese character mistake of identification unavoidably; One is that mandarin standard pronunciation phonetically similar word mistake occurs when converting the Chinese phonetic alphabet to;
When phonetic entry generation Chinese Character Recognition mistake, following method can be adopted to revise:
Method one, statement displacement amendment: by the sequence label of word before erroneous words, screen on all correct word allowing before erroneous words, mark sequence reorientated in remaining Chinese character, and meanwhile, cursor highlights and is positioned in erroneous words, choose correct candidate word and substitute erroneous words, or re-enter correct voice;
1. sequence is designated as, 2., 3., 4., 5., 6., 7., 8., 9.,
chinese character, screen on the numerical key on direct keypad; Sequence is designated as
chinese character, by [0+ numeral] upper screen;
Method two, cursor location amendment: mobile cursor, navigates to the amendment of erroneous words place, and when cursor location Chinese character, Chinese character can highlight display;
Method three, sequence demarcate position amendment: 1. sequence is designated as, 2., 3., 4., 5., 6., 7., 8., 9.,
chinese character, by [Ctrl+ numeral], a key quick position is to erroneous words place; Sequence is designated as
chinese character, by [Alt+ numeral], a key quick position is to erroneous words place;
Adopt the amendment of cursor location and sequence to demarcate position amendment, cursor is in location and also locate while highlighting Chinese character and highlight pinyin syllable corresponding to Chinese character.
2. a kind of Chinese speech input method as claimed in claim 1 and statement are located and error correction method, it is characterized in that: do not limit to the sequence reference symbol number of Chinese character and syllable location, sequence target position does not limit:
A, be used for Chinese character and syllable location mark sequence sequence mark, be not limited to 1., 2., 3., 4., 5., 6., 7., 8., 9.,
sequence, can adopt other digital forms, letter or symbol combination mark sequence, and 1,2,3,4,5,6,7,8,9,0 also can be adopted to assign to mark sequence with the chromatic zones that front and back group is different simultaneously;
B, sequence target position, be not limited to immediately below Chinese character and syllable, also can directly over Chinese character and syllable, or subscript or subscript mode.
3. a kind of Chinese speech input method as claimed in claim 1 and statement are located and error correction method, it is characterized in that: the candidate Chinese character layout style of show candidate words does not limit, can be criss-cross, horizontal stripe type, shape, vertical shaped as frame or other forms of design style; The relative position of candidate's space of a whole page and statement line does not limit, and can be positioned at above statement line or below.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510557031.1A CN105117034A (en) | 2015-08-31 | 2015-08-31 | Method for inputting Chinese speeches, positioning statements and correcting errors |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510557031.1A CN105117034A (en) | 2015-08-31 | 2015-08-31 | Method for inputting Chinese speeches, positioning statements and correcting errors |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105117034A true CN105117034A (en) | 2015-12-02 |
Family
ID=54665049
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510557031.1A Pending CN105117034A (en) | 2015-08-31 | 2015-08-31 | Method for inputting Chinese speeches, positioning statements and correcting errors |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105117034A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105513586A (en) * | 2015-12-18 | 2016-04-20 | 百度在线网络技术（北京）有限公司 | Speech recognition result display method and speech recognition result display device |
| CN106406807A (en) * | 2016-09-19 | 2017-02-15 | 北京云知声信息技术有限公司 | A method and a device for voice correction of characters |
| CN106782543A (en) * | 2017-03-24 | 2017-05-31 | 联想(北京)有限公司 | A kind of information processing method and electronic equipment |
| CN107992457A (en) * | 2017-12-01 | 2018-05-04 | 深圳乐信软件技术有限公司 | A kind of information conversion method, device, terminal device and storage medium |
| CN109215660A (en) * | 2018-07-09 | 2019-01-15 | 维沃移动通信有限公司 | Text error correction method and mobile terminal after speech recognition |
| CN111028834A (en) * | 2019-10-30 | 2020-04-17 | 支付宝(杭州)信息技术有限公司 | Voice message reminding method and device, server and voice message reminding equipment |
| CN112270930A (en) * | 2020-10-22 | 2021-01-26 | 江苏峰鑫网络科技有限公司 | Method for voice recognition conversion |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1412741A (en) * | 2002-12-13 | 2003-04-23 | 郑方 | Chinese speech identification method with dialect background |
| CN101118541A (en) * | 2006-08-03 | 2008-02-06 | 苗玉水 | Chinese-voice-code voice recognizing method |
| CN101710262A (en) * | 2009-12-11 | 2010-05-19 | 北京搜狗科技发展有限公司 | Error correction method and error correction device of characters |
| CN103164184A (en) * | 2013-03-22 | 2013-06-19 | 祁磊 | Real-time recording device for court criminal justice |
| CN103207857A (en) * | 2013-02-27 | 2013-07-17 | 何玉春 | Intelligent language management system |
| CN103578465A (en) * | 2013-10-18 | 2014-02-12 | 威盛电子股份有限公司 | Speech recognition method and electronic device |
| CN104049770A (en) * | 2014-07-03 | 2014-09-17 | 任文 | Chinese character candidate selecting, locating and correcting scheme based on Chinese sentence input |
-
2015
- 2015-08-31 CN CN201510557031.1A patent/CN105117034A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1412741A (en) * | 2002-12-13 | 2003-04-23 | 郑方 | Chinese speech identification method with dialect background |
| CN101118541A (en) * | 2006-08-03 | 2008-02-06 | 苗玉水 | Chinese-voice-code voice recognizing method |
| CN101710262A (en) * | 2009-12-11 | 2010-05-19 | 北京搜狗科技发展有限公司 | Error correction method and error correction device of characters |
| CN103207857A (en) * | 2013-02-27 | 2013-07-17 | 何玉春 | Intelligent language management system |
| CN103164184A (en) * | 2013-03-22 | 2013-06-19 | 祁磊 | Real-time recording device for court criminal justice |
| CN103578465A (en) * | 2013-10-18 | 2014-02-12 | 威盛电子股份有限公司 | Speech recognition method and electronic device |
| CN104049770A (en) * | 2014-07-03 | 2014-09-17 | 任文 | Chinese character candidate selecting, locating and correcting scheme based on Chinese sentence input |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105513586A (en) * | 2015-12-18 | 2016-04-20 | 百度在线网络技术（北京）有限公司 | Speech recognition result display method and speech recognition result display device |
| CN106406807A (en) * | 2016-09-19 | 2017-02-15 | 北京云知声信息技术有限公司 | A method and a device for voice correction of characters |
| CN106782543A (en) * | 2017-03-24 | 2017-05-31 | 联想(北京)有限公司 | A kind of information processing method and electronic equipment |
| CN107992457A (en) * | 2017-12-01 | 2018-05-04 | 深圳乐信软件技术有限公司 | A kind of information conversion method, device, terminal device and storage medium |
| CN109215660A (en) * | 2018-07-09 | 2019-01-15 | 维沃移动通信有限公司 | Text error correction method and mobile terminal after speech recognition |
| CN111028834A (en) * | 2019-10-30 | 2020-04-17 | 支付宝(杭州)信息技术有限公司 | Voice message reminding method and device, server and voice message reminding equipment |
| CN111028834B (en) * | 2019-10-30 | 2023-01-20 | 蚂蚁财富(上海)金融信息服务有限公司 | Voice message reminding method and device, server and voice message reminding equipment |
| CN112270930A (en) * | 2020-10-22 | 2021-01-26 | 江苏峰鑫网络科技有限公司 | Method for voice recognition conversion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105117034A (en) | Method for inputting Chinese speeches, positioning statements and correcting errors | |
| US8504350B2 (en) | User-interactive automatic translation device and method for mobile device | |
| US9292499B2 (en) | Automatic translation and interpretation apparatus and method | |
| TWI293455B (en) | System and method for disambiguating phonetic input | |
| CN104166462B (en) | The input method and system of a kind of word | |
| JP4829901B2 (en) | Method and apparatus for confirming manually entered indeterminate text input using speech input | |
| US20170270086A1 (en) | Apparatus, method, and computer program product for correcting speech recognition error | |
| CN105404621B (en) | A kind of method and system that Chinese character is read for blind person | |
| US20080180283A1 (en) | System and method of cross media input for chinese character input in electronic equipment | |
| JP5535238B2 (en) | Information processing device | |
| CN101196792A (en) | Automatic correction method and device for document file | |
| US20150081270A1 (en) | Speech translation apparatus, speech translation method, and non-transitory computer readable medium thereof | |
| CN101133411A (en) | Fault-tolerant romanized input method for non-roman characters | |
| CN103578464A (en) | Language model establishing method, speech recognition method and electronic device | |
| US20070288240A1 (en) | User interface for text-to-phone conversion and method for correcting the same | |
| US20070040707A1 (en) | Separation of Components and Characters in Chinese Text Input | |
| US20120109633A1 (en) | Method and system for diacritizing arabic language text | |
| CN106802722A (en) | A kind of pronunciation inputting method and system based on three-stroke digital input method | |
| JP2007317163A (en) | Electronic dictionary and retrieval method | |
| JP2004271895A (en) | Multilingual speech recognition system and pronunciation learning system | |
| KR20230081248A (en) | System and method for context-based speech recognition | |
| CN114528826A (en) | Dictionary editing device, dictionary editing method, and recording medium having dictionary editing program recorded therein | |
| Thangthai et al. | Automatic syllable-pattern induction in statistical Thai text-to-phone transcription. | |
| KR20160054751A (en) | System for editing a text and method thereof | |
| US11544301B2 (en) | Identification method with multi-type input and electronic device using the same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20151202 |