CN106098078A - A kind of audio recognition method that may filter that speaker noise and system thereof - Google Patents
A kind of audio recognition method that may filter that speaker noise and system thereof Download PDFInfo
- Publication number
- CN106098078A CN106098078A CN201610413367.5A CN201610413367A CN106098078A CN 106098078 A CN106098078 A CN 106098078A CN 201610413367 A CN201610413367 A CN 201610413367A CN 106098078 A CN106098078 A CN 106098078A
- Authority
- CN
- China
- Prior art keywords
- frequency
- synthesized voice
- speech
- user speech
- amplitude
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000005070 sampling Methods 0.000 claims abstract description 7
- 239000000203 mixture Substances 0.000 claims abstract description 6
- 238000006243 chemical reaction Methods 0.000 claims description 12
- 238000001914 filtration Methods 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 7
- 230000015572 biosynthetic process Effects 0.000 claims description 6
- 238000003786 synthesis reaction Methods 0.000 claims description 6
- 238000011084 recovery Methods 0.000 claims description 4
- 230000007613 environmental effect Effects 0.000 abstract description 5
- 238000005516 engineering process Methods 0.000 description 6
- 239000002131 composite material Substances 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 241000238558 Eucarida Species 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
- G10L21/0216—Noise filtering characterised by the method used for estimating noise
- G10L21/0232—Processing in the frequency domain
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; 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 TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0316—Speech enhancement, e.g. noise reduction or echo cancellation by changing the amplitude
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Computational Linguistics (AREA)
- Quality & Reliability (AREA)
- Signal Processing (AREA)
- Telephone Function (AREA)
- Circuit For Audible Band Transducer (AREA)
- Soundproofing, Sound Blocking, And Sound Damping (AREA)
Abstract
The invention provides a kind of audio recognition method that may filter that speaker noise and system thereof, method includes: when detecting by mike typing user speech and speaker storaged voice file in playing intelligent terminal being detected, then obtain user speech and the synthesized voice of loudspeaker sound；According to the first frequency of loudspeaker sound of sampling in intelligent terminal and the first amplitude, and the synthesized voice frequency of synthesized voice and synthesized voice amplitude, it is calculated second frequency and second amplitude of user speech；Filter the tone color of loudspeaker sound in synthesized voice, and restore obtain user speech with second frequency and second amplitude of user speech；According to speech database, user speech is converted into text.Present invention achieves user is using speech recognition software and speaker when playing outer sound, terminal inner treater is analyzed according to sound composition, filter out loudspeaker sound so that the user speech that backstage receives reduces environmental noise, it is achieved the efficient identification of voice.
Description
Technical field
The present invention relates to technical field of voice recognition, particularly relate to a kind of audio recognition method that may filter that speaker noise
And system.
Background technology
Speech recognition technology the most progressively becomes the key technology of man-machine interface, speech recognition technology and language in information technology
Sound synthetic technology combines and enables people to get rid of keyboard, is operated by voice command.The rise of mobile Internet just becomes
The most important applied environment of speech recognition, such as the Siri of Apple, domestic news fly software etc., it is possible to identify user efficiently
Voice.At present intelligent terminal is upper can install similar software, it is possible to user speech is converted into word, and by voice and after
Platform data base mate, and generates text importing, is the most directly controlled.In order to efficient identification voice, need user
Avoid environmental noise during input voice as far as possible.
But, when intelligent terminal is when playing music, and user speaks facing to mike, can bring the musical sound of speaker into,
Recognition efficiency is caused to decline to a great extent.
Therefore, prior art could be improved and develop.
Summary of the invention
In place of above-mentioned the deficiencies in the prior art, it is an object of the invention to provide a kind of speaker noise of may filter that
Audio recognition method and system thereof, it is intended in solution prior art, intelligent terminal is when playing music, and user says facing to mike
Words, can bring the musical sound of speaker into, cause the problem that recognition efficiency declines to a great extent.
In order to achieve the above object, this invention takes techniques below scheme:
A kind of audio recognition method that may filter that speaker noise, wherein, said method comprising the steps of:
A, when detecting by mike typing user speech and detecting that speaker is playing storaged voice literary composition in intelligent terminal
During part, then obtain user speech and the synthesized voice of loudspeaker sound；
B, according to the first frequency of loudspeaker sound of sampling in intelligent terminal and the first amplitude, and the synthesis of described synthesized voice
Voice frequency and synthesized voice amplitude, be calculated second frequency and second amplitude of user speech；
C, filter the tone color of loudspeaker sound in described synthesized voice, and restore with second frequency and second amplitude of user speech
Obtain user speech；
D, according to speech database, user speech is converted into text.
The described audio recognition method that may filter that speaker noise, wherein, described step B specifically includes:
B1, according to the least common multiple that synthesized voice frequency is first frequency and second frequency, by synthesized voice frequency and first frequency
It is calculated second frequency；
B2, according to synthesized voice amplitude and the difference of the first amplitude, be calculated the second amplitude.
The described audio recognition method that may filter that speaker noise, wherein, described step C specifically includes:
C1, by synthesized voice by after audio coder analog/digital conversion, will have synthesized voice frequency, synthesized voice amplitude and synthesized voice
The synthesized voice coding of tone color delivers to processor；
C2, processor filter out the tone color of loudspeaker sound in described synthesized voice, retain the tone color of user speech；
The second frequency of user speech and the second amplitude are changed into part of speech by C3, audio decoder, described part of speech with
The tone color of user speech is restored and is obtained user speech.
The described audio recognition method that may filter that speaker noise, wherein, described step D specifically includes:
D1, user speech is uploaded to the speech database in high in the clouds；
D2, user speech is mated in speech database, obtain text；
D3, described text is sent to intelligent terminal, and show.
The described audio recognition method that may filter that speaker noise, wherein, also includes in described step A that processor obtains sound
Frequently the loudspeaker sound coding of each frame of loudspeaker sound in encoder.
A kind of speech recognition system that may filter that speaker noise, wherein, including:
Detection and acquisition module, for when detecting by mike typing user speech and detecting that speaker is playing intelligence
In energy terminal during storaged voice file, then obtain user speech and the synthesized voice of loudspeaker sound；
Computing module, for the first frequency according to the loudspeaker sound sampled in intelligent terminal and the first amplitude, and described conjunction
The synthesized voice frequency of audio and synthesized voice amplitude, be calculated second frequency and second amplitude of user speech；
Filter and restoration module, for filtering the tone color of loudspeaker sound in described synthesized voice, and with the second frequency of user speech
Rate and the second amplitude restore and obtain user speech；
Conversion module, for according to speech database, is converted into text by user speech.
The described speech recognition system that may filter that speaker noise, wherein, described computing module specifically includes:
Frequency computing unit, for according to the least common multiple that synthesized voice frequency is first frequency and second frequency, by synthesized voice
Frequency and first frequency, be calculated second frequency；
Magnitude determinations unit, according to synthesized voice amplitude and the difference of the first amplitude, is calculated the second amplitude.
The described speech recognition system that may filter that speaker noise, wherein, described filtration and restoration module specifically include:
Coding transmitting element, after by synthesized voice by audio coder analog/digital conversion, will have synthesized voice frequency, synthesis
The synthesized voice coding of sound amplitude and synthesized voice tone color delivers to processor；
Filter element, processor filters out the tone color of loudspeaker sound in described synthesized voice, retains the tone color of user speech；
Restoration unit, second frequency and second amplitude of user speech are changed into part of speech, described part by audio decoder
Voice obtains user speech with the tone color recovery of user speech.
The described speech recognition system that may filter that speaker noise, wherein, described conversion module specifically includes:
Uploading unit, for being uploaded to the speech database in high in the clouds by user speech；
Matching unit, for being mated in speech database by user speech, obtains text；
Send display unit, for described text is sent to intelligent terminal, and show.
The described speech recognition system that may filter that speaker noise, wherein, is additionally operable to place in described detection and acquisition module
Reason device obtains the loudspeaker sound coding of each frame of loudspeaker sound in audio coder.
The audio recognition method that may filter that speaker noise of the present invention and system thereof, method includes: when detecting
By mike typing user speech and detect that speaker when playing storaged voice file in intelligent terminal, then obtains use
Family voice and the synthesized voice of loudspeaker sound；Shake according to the first frequency and first of the loudspeaker sound of sampling in intelligent terminal
Width, and the synthesized voice frequency of synthesized voice and synthesized voice amplitude, be calculated second frequency and second amplitude of user speech；Cross
The tone color of loudspeaker sound in filter synthesized voice, and restore obtain user speech with second frequency and second amplitude of user speech；
According to speech database, user speech is converted into text.Present invention achieves user using speech recognition software and raising
Sound device is when playing outer sound, and the processor in terminal is analyzed according to the composition of sound, filters out loudspeaker sound so that after
The user speech that platform receives reduces environmental noise, it is achieved the efficient identification of voice.
Accompanying drawing explanation
Fig. 1 is the flow chart of the audio recognition method preferred embodiment that may filter that speaker noise of the present invention.
Fig. 2 is acquisition user speech in the audio recognition method preferred embodiment that may filter that speaker noise of the present invention
Second frequency and the particular flow sheet of the second amplitude.
Fig. 3 is to restore in the audio recognition method preferred embodiment that may filter that speaker noise of the present invention to obtain user
The particular flow sheet of voice.
Fig. 4 is the tool converting text in the audio recognition method preferred embodiment that may filter that speaker noise of the present invention
Body flow chart.
Fig. 5 is the structured flowchart of the speech recognition system preferred embodiment that may filter that speaker noise of the present invention.
Detailed description of the invention
The present invention provides a kind of audio recognition method that may filter that speaker noise and system thereof, for making the mesh of the present invention
, technical scheme and effect clearer, clear and definite, the present invention is described in more detail for the embodiment that develops simultaneously referring to the drawings.
Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
Refer to Fig. 1, it is the stream of the audio recognition method preferred embodiment that may filter that speaker noise of the present invention
Cheng Tu.May filter that the audio recognition method of speaker noise described in as it is shown in figure 1, comprise the following steps:
Step S100, when detecting by mike typing user speech and detecting that speaker is deposited playing in intelligent terminal
During storage voice document, then obtain user speech and the synthesized voice of loudspeaker sound.
In the present embodiment, when player during user opens intelligent terminal, it is possible to synchronize to open the speech recognition on backstage
Process, such intelligent terminal can detect user's whether typing voice in real time when playing music.Once detect intelligent terminal
On by player plays voice document, and when having user voice typing, then obtain user speech and the synthesis of loudspeaker sound
Sound.Now, during without any process, user speech and loudspeaker sound also cannot be distinguished by out, and this is accomplished by the place of subsequent step
Reason.
Step S200, according to the first frequency of loudspeaker sound of sampling in intelligent terminal and the first amplitude, and described conjunction
The synthesized voice frequency of audio and synthesized voice amplitude, be calculated second frequency and second amplitude of user speech.
In the present embodiment, owing to speaker material and structure are fixing, therefore speaker tone color is in intelligent terminal
Processor be known.Same, player is during playing voice document, and processor obtains in audio coder and raises
The loudspeaker sound coding of each frame of sound device sound, can obtain each frame of voice data in loudspeaker sound by player
First frequency and the first amplitude.
Since it is known that the first frequency of loudspeaker sound and the first amplitude, and the Composite tone of described synthesized voice
Rate and synthesized voice amplitude, therefore can be tried to achieve according to the least common multiple that synthesized voice frequency is first frequency and second frequency
Two frequencies, it is possible to be the first amplitude according to synthesized voice amplitude and the second amplitude sum tries to achieve the second amplitude.So, by processing
Device simply calculating processes, and i.e. can get second frequency and second amplitude of user speech.
Step S300, filter the tone color of loudspeaker sound in described synthesized voice, and with the second frequency and of user speech
Two amplitudes restore and obtain user speech.
When after the second frequency obtaining user speech and the second amplitude, owing to optionally filtering out speaker
(owing to material and the structure of speaker are fixing, therefore speaker tone color is for the processor in intelligent terminal to tone color
Know), only retain the tone color of user speech, so obtain by the tone color of user speech, second frequency and the second amplitude are resilient
To user speech.So, filtered out loudspeaker sound part in synthesized voice, only remained the part of user speech, it is achieved that
The speech recognition effect of filtering speaker noise.
Step S400, according to speech database, user speech is converted into text.
After user speech is mated by speech database, then transfer the text for correspondence to, according to the finger corresponding to text
Order carries out the operation of correspondence to intelligent terminal.Such as, during user opens player plays music, the voice on backstage is known
Other process detection is to user's typing voice " F.F. 10 seconds ", then, after passing through the process of step S100-S400, be converted into text " fast
Enter 10 seconds ".Now, player according to control instruction F.F. corresponding to the text by currently playing voice document F.F. 10 seconds.This
Sample achieves in the case of sound of having powerful connections, the accurate identification to user speech.
Further, as in figure 2 it is shown, in the described audio recognition method that may filter that speaker noise, described step
S200 specifically includes:
Step S201, according to the least common multiple that synthesized voice frequency is first frequency and second frequency, by synthesized voice frequency and
One frequency is calculated second frequency.
Due to after loudspeaker sound and user speech form synthesized voice, processor is the Composite tone of synthesized voice of can sampling
Rate and synthesized voice amplitude.It is furthermore also known that the least common multiple that synthesized voice frequency is first frequency and second frequency, i.e. 1/ closes
Audio frequency=N* (1/ first frequency) * (1/ second frequency), wherein N is any positive integer.According to above formula, can solve and obtain
Two frequencies.
Step S202, according to synthesized voice amplitude and the difference of the first amplitude, be calculated the second amplitude.
Further, as it is shown on figure 3, in the described audio recognition method that may filter that speaker noise, described step
S300 specifically includes:
Step S301, by synthesized voice by after audio coder analog/digital conversion, will have synthesized voice frequency, synthesized voice amplitude and
The synthesized voice coding of synthesized voice tone color delivers to processor；
Step S302, processor filter out the tone color of loudspeaker sound in described synthesized voice, retain the tone color of user speech；
Second frequency and second amplitude of user speech are changed into part of speech, described part by step S303, audio decoder
Voice obtains user speech with the tone color recovery of user speech.
Further, as shown in Figure 4, in the described audio recognition method that may filter that speaker noise, described step
S400 specifically includes:
Step S401, user speech is uploaded to the speech database in high in the clouds；
Step S402, user speech is mated in speech database, obtain text；
Step S403, described text is sent to intelligent terminal, and show.
Visible, present invention achieves user use speech recognition software and speaker when playing outer sound, in terminal
Processor be analyzed according to the composition of sound, filter out loudspeaker sound so that backstage receive user speech in reduce
Environmental noise, it is achieved the efficient identification of voice.
Based on said method embodiment, present invention also offers a kind of speech recognition system that may filter that speaker noise.
The speech recognition system of speaker noise is may filter that described in as it is shown in figure 5, including:
Detection and acquisition module 100, for when detecting by mike typing user speech and detecting that speaker is being play
In intelligent terminal during storaged voice file, then obtain user speech and the synthesized voice of loudspeaker sound；
Computing module 200, for the first frequency according to the loudspeaker sound sampled in intelligent terminal and the first amplitude, and institute
State synthesized voice frequency and the synthesized voice amplitude of synthesized voice, be calculated second frequency and second amplitude of user speech；
Filter and restoration module 300, for filtering the tone color of loudspeaker sound in described synthesized voice, and with the second of user speech
Frequency and the second amplitude restore and obtain user speech；
Conversion module 400, for according to speech database, is converted into text by user speech.
Further, in the described speech recognition system that may filter that speaker noise, described computing module 200 specifically wraps
Include:
Frequency computing unit, for according to the least common multiple that synthesized voice frequency is first frequency and second frequency, by synthesized voice
Frequency and first frequency are calculated second frequency；
Magnitude determinations unit, according to synthesized voice amplitude and the difference of the first amplitude, is calculated the second amplitude.
Further, in the described speech recognition system that may filter that speaker noise, described filtration and restoration module 300
Specifically include:
Coding transmitting element, after by synthesized voice by audio coder analog/digital conversion, will have synthesized voice frequency, synthesis
The synthesized voice coding of sound amplitude and synthesized voice tone color delivers to processor；
Filter element, processor filters out the tone color of loudspeaker sound in described synthesized voice, retains the tone color of user speech；
Restoration unit, second frequency and second amplitude of user speech are changed into part of speech, described part by audio decoder
Voice obtains user speech with the tone color recovery of user speech.
Further, in the described speech recognition system that may filter that speaker noise, described conversion module 400 specifically wraps
Include:
Uploading unit, for being uploaded to the speech database in high in the clouds by user speech；
Matching unit, for being mated in speech database by user speech, obtains text；
Send display unit, for described text is sent to intelligent terminal, and show.
Further, in the described speech recognition system that may filter that speaker noise, described detection and acquisition module 100
It is additionally operable to processor and obtains the loudspeaker sound coding of each frame of loudspeaker sound in audio coder.
In sum, the audio recognition method that may filter that speaker noise of the present invention and system thereof, method includes:
When detecting by mike typing user speech and speaker storaged voice file in playing intelligent terminal being detected,
Then obtain user speech and the synthesized voice of loudspeaker sound；According in intelligent terminal sampling loudspeaker sound first frequency,
And first amplitude, and the synthesized voice frequency of synthesized voice and synthesized voice amplitude, it is calculated the second frequency and of user speech
Two amplitudes；Filter the tone color of loudspeaker sound in synthesized voice, and restore obtain with second frequency and second amplitude of user speech
User speech；According to speech database, user speech is converted into text.Present invention achieves user is using speech recognition soft
Part and speaker are when playing outer sound, and the processor in terminal is analyzed according to the composition of sound, filters out speaker sound
Sound so that reduce environmental noise in the user speech that backstage receives, it is achieved the efficient identification of voice.
It is understood that for those of ordinary skills, can according to technical scheme and this
Bright design in addition equivalent or change, and all these change or replace the guarantor that all should belong to appended claims of the invention
Protect scope.
Claims (10)
1. the audio recognition method that may filter that speaker noise, it is characterised in that said method comprising the steps of:
A, when detecting by mike typing user speech and detecting that speaker is playing storaged voice literary composition in intelligent terminal
During part, then obtain user speech and the synthesized voice of loudspeaker sound；
B, according to the first frequency of loudspeaker sound of sampling in intelligent terminal and the first amplitude, and the synthesis of described synthesized voice
Voice frequency and synthesized voice amplitude, be calculated second frequency and second amplitude of user speech；
C, filter the tone color of loudspeaker sound in described synthesized voice, and restore with second frequency and second amplitude of user speech
Obtain user speech；
D, according to speech database, user speech is converted into text.
May filter that the audio recognition method of speaker noise the most according to claim 1, it is characterised in that described step B has
Body includes:
B1, according to the least common multiple that synthesized voice frequency is first frequency and second frequency, by synthesized voice frequency and first frequency
It is calculated second frequency；
B2, according to synthesized voice amplitude and the difference of the first amplitude, be calculated the second amplitude.
May filter that the audio recognition method of speaker noise the most according to claim 1, it is characterised in that described step C has
Body includes:
C1, by synthesized voice by after audio coder analog/digital conversion, will have synthesized voice frequency, synthesized voice amplitude and synthesized voice
The synthesized voice coding of tone color delivers to processor；
C2, processor filter out the tone color of loudspeaker sound in described synthesized voice, retain the tone color of user speech；
The second frequency of user speech and the second amplitude are changed into part of speech by C3, audio decoder, described part of speech with
The tone color of user speech is restored and is obtained user speech.
May filter that the audio recognition method of speaker noise the most according to claim 1, it is characterised in that described step D has
Body includes:
D1, user speech is uploaded to the speech database in high in the clouds；
D2, user speech is mated in speech database, obtain text；
D3, described text is sent to intelligent terminal, and show.
May filter that the audio recognition method of speaker noise the most according to claim 1, it is characterised in that in described step A
Also include that processor obtains the loudspeaker sound coding of each frame of loudspeaker sound in audio coder.
6. the speech recognition system that may filter that speaker noise, it is characterised in that including:
Detection and acquisition module, for when detecting by mike typing user speech and detecting that speaker is playing intelligence
In energy terminal during storaged voice file, then obtain user speech and the synthesized voice of loudspeaker sound；
Computing module, for the first frequency according to the loudspeaker sound sampled in intelligent terminal and the first amplitude, and described conjunction
The synthesized voice frequency of audio and synthesized voice amplitude, be calculated second frequency and second amplitude of user speech；
Filter and restoration module, for filtering the tone color of loudspeaker sound in described synthesized voice, and with the second frequency of user speech
Rate and the second amplitude restore and obtain user speech；
Conversion module, for according to speech database, is converted into text by user speech.
May filter that the speech recognition system of speaker noise the most according to claim 6, it is characterised in that described computing module
Specifically include:
Frequency computing unit, for according to the least common multiple that synthesized voice frequency is first frequency and second frequency, by synthesized voice
Frequency and first frequency are calculated second frequency；
Magnitude determinations unit, according to synthesized voice amplitude and the difference of the first amplitude, is calculated the second amplitude.
May filter that the speech recognition system of speaker noise the most according to claim 6, it is characterised in that described filtration and multiple
Grand master pattern block specifically includes:
Coding transmitting element, after by synthesized voice by audio coder analog/digital conversion, will have synthesized voice frequency, synthesis
The synthesized voice coding of sound amplitude and synthesized voice tone color delivers to processor；
Filter element, processor filters out the tone color of loudspeaker sound in described synthesized voice, retains the tone color of user speech；
Restoration unit, second frequency and second amplitude of user speech are changed into part of speech, described part by audio decoder
Voice obtains user speech with the tone color recovery of user speech.
May filter that the speech recognition system of speaker noise the most according to claim 6, it is characterised in that described conversion module
Specifically include:
Uploading unit, for being uploaded to the speech database in high in the clouds by user speech；
Matching unit, for being mated in speech database by user speech, obtains text；
Send display unit, for described text is sent to intelligent terminal, and show.
May filter that the speech recognition system of speaker noise the most according to claim 6, it is characterised in that described detection and
Acquisition module is additionally operable to processor and obtains the loudspeaker sound coding of each frame of loudspeaker sound in audio coder.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610413367.5A CN106098078B (en) | 2016-06-14 | 2016-06-14 | Voice recognition method and system capable of filtering loudspeaker noise |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610413367.5A CN106098078B (en) | 2016-06-14 | 2016-06-14 | Voice recognition method and system capable of filtering loudspeaker noise |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106098078A true CN106098078A (en) | 2016-11-09 |
| CN106098078B CN106098078B (en) | 2020-06-02 |
Family
ID=57845701
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610413367.5A Active CN106098078B (en) | 2016-06-14 | 2016-06-14 | Voice recognition method and system capable of filtering loudspeaker noise |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106098078B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106569774A (en) * | 2016-11-11 | 2017-04-19 | 青岛海信移动通信技术股份有限公司 | Method of eliminating noise, and terminal |
| CN108173740A (en) * | 2017-11-30 | 2018-06-15 | 维沃移动通信有限公司 | A kind of method and apparatus of voice communication |
| CN108335701A (en) * | 2018-01-24 | 2018-07-27 | 青岛海信移动通信技术股份有限公司 | A kind of method and apparatus carrying out noise reduction |
| CN110164432A (en) * | 2019-03-26 | 2019-08-23 | 北京海益同展信息科技有限公司 | A kind of Internet data center's method for inspecting and device |
| CN110797048A (en) * | 2018-08-01 | 2020-02-14 | 珠海格力电器股份有限公司 | Method and device for acquiring voice information |
| CN111583495A (en) * | 2020-04-30 | 2020-08-25 | 厦门快商通科技股份有限公司 | Remote access control system based on voice recognition and authorization method thereof |
| CN111583496A (en) * | 2020-04-30 | 2020-08-25 | 厦门快商通科技股份有限公司 | Remote access control system based on voice recognition and control method thereof |
| 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 |
| CN112887856A (en) * | 2021-01-25 | 2021-06-01 | 湖南普奇水环境研究院有限公司 | Sound processing method and system for reducing noise |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102655006A (en) * | 2011-03-03 | 2012-09-05 | 富泰华工业（深圳）有限公司 | Voice transmission device and voice transmission method |
| KR101396873B1 (en) * | 2013-04-03 | 2014-05-19 | 주식회사 크린컴 | Method and apparatus for noise reduction in a communication device having two microphones |
| CN105516859A (en) * | 2015-11-27 | 2016-04-20 | 深圳Tcl数字技术有限公司 | Method and system for eliminating echo |
| CN105657150A (en) * | 2015-09-29 | 2016-06-08 | 宇龙计算机通信科技(深圳)有限公司 | Noise elimination method and device and electronic device |
-
2016
- 2016-06-14 CN CN201610413367.5A patent/CN106098078B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102655006A (en) * | 2011-03-03 | 2012-09-05 | 富泰华工业（深圳）有限公司 | Voice transmission device and voice transmission method |
| KR101396873B1 (en) * | 2013-04-03 | 2014-05-19 | 주식회사 크린컴 | Method and apparatus for noise reduction in a communication device having two microphones |
| CN105657150A (en) * | 2015-09-29 | 2016-06-08 | 宇龙计算机通信科技(深圳)有限公司 | Noise elimination method and device and electronic device |
| CN105516859A (en) * | 2015-11-27 | 2016-04-20 | 深圳Tcl数字技术有限公司 | Method and system for eliminating echo |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106569774A (en) * | 2016-11-11 | 2017-04-19 | 青岛海信移动通信技术股份有限公司 | Method of eliminating noise, and terminal |
| CN106569774B (en) * | 2016-11-11 | 2020-07-10 | 青岛海信移动通信技术股份有限公司 | Method and terminal for removing noise |
| CN108173740A (en) * | 2017-11-30 | 2018-06-15 | 维沃移动通信有限公司 | A kind of method and apparatus of voice communication |
| CN108335701A (en) * | 2018-01-24 | 2018-07-27 | 青岛海信移动通信技术股份有限公司 | A kind of method and apparatus carrying out noise reduction |
| CN108335701B (en) * | 2018-01-24 | 2021-04-13 | 青岛海信移动通信技术股份有限公司 | Method and equipment for sound noise reduction |
| CN110797048A (en) * | 2018-08-01 | 2020-02-14 | 珠海格力电器股份有限公司 | Method and device for acquiring voice information |
| CN110164432A (en) * | 2019-03-26 | 2019-08-23 | 北京海益同展信息科技有限公司 | A kind of Internet data center's method for inspecting and device |
| CN111583495A (en) * | 2020-04-30 | 2020-08-25 | 厦门快商通科技股份有限公司 | Remote access control system based on voice recognition and authorization method thereof |
| CN111583496A (en) * | 2020-04-30 | 2020-08-25 | 厦门快商通科技股份有限公司 | Remote access control system based on voice recognition and control method thereof |
| 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 |
| CN112887856A (en) * | 2021-01-25 | 2021-06-01 | 湖南普奇水环境研究院有限公司 | Sound processing method and system for reducing noise |
| CN112887856B (en) * | 2021-01-25 | 2023-03-24 | 湖南普奇水环境研究院有限公司 | Sound processing method and system for reducing noise |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106098078B (en) | 2020-06-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106098078A (en) | A kind of audio recognition method that may filter that speaker noise and system thereof | |
| US9978388B2 (en) | Systems and methods for restoration of speech components | |
| RU2373584C2 (en) | Method and device for increasing speech intelligibility using several sensors | |
| KR102225404B1 (en) | Method and Apparatus of Speech Recognition Using Device Information | |
| KR20060044629A (en) | Isolating speech signals utilizing neural networks | |
| CN103377651B (en) | The automatic synthesizer of voice and method | |
| CN108461081B (en) | Voice control method, device, equipment and storage medium | |
| MX2007015446A (en) | Multi-sensory speech enhancement using a speech-state model. | |
| CN101510905A (en) | Method and apparatus for multi-sensory speech enhancement on a mobile device | |
| KR20130033372A (en) | Speech audio processing | |
| CN103886863A (en) | Audio processing device and audio processing method | |
| CN113129867B (en) | Training method of voice recognition model, voice recognition method, device and equipment | |
| CN115602165B (en) | Digital employee intelligent system based on financial system | |
| CN111883135A (en) | Voice transcription method and device and electronic equipment | |
| CN103474062A (en) | Voice identification method | |
| Jaroslavceva et al. | Robot Ego‐Noise Suppression with Labanotation‐Template Subtraction | |
| CN106228984A (en) | Voice recognition information acquisition methods | |
| Yu et al. | Text-Dependent Speech Enhancement for Small-Footprint Robust Keyword Detection. | |
| Principi et al. | A speech-based system for in-home emergency detection and remote assistance | |
| Balasubramanian et al. | Estimation of ideal binary mask for audio-visual monaural speech enhancement | |
| Zhu | [Retracted] Multimedia Recognition of Piano Music Based on the Hidden Markov Model | |
| Zheng et al. | Bandwidth extension WaveNet for bone-conducted speech enhancement | |
| CN104078049B (en) | Signal processing apparatus and signal processing method | |
| Vicente-Peña et al. | Band-pass filtering of the time sequences of spectral parameters for robust wireless speech recognition | |
| CN114333892A (en) | Voice processing method and device, electronic equipment and readable medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |