CN108154880A - The robot that environmental noise carries out speech recognition can be differentiated in real time - Google Patents
The robot that environmental noise carries out speech recognition can be differentiated in real time Download PDFInfo
- Publication number
- CN108154880A CN108154880A CN201611103797.3A CN201611103797A CN108154880A CN 108154880 A CN108154880 A CN 108154880A CN 201611103797 A CN201611103797 A CN 201611103797A CN 108154880 A CN108154880 A CN 108154880A
- Authority
- CN
- China
- Prior art keywords
- peak value
- environmental noise
- robot
- audio data
- threshold values
- 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
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
- G10L15/00—Speech recognition
- G10L15/20—Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech
-
- 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
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/78—Detection of presence or absence of voice signals
- G10L25/84—Detection of presence or absence of voice signals for discriminating voice from noise
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Manipulator (AREA)
Abstract
The invention discloses a kind of robot that can be differentiated environmental noise in real time and carry out speech recognition, it is characterized in that：The robot analyzes the peak value that volume is represented in audio data by real-time sampling, and peak value for a long time, regular is included by environmental noise threshold values；The robot by real-time sampling analyze audio data in occur new long-time, regularity existing for peak value when, new peak value is updated to new environmental noise threshold values.It is found when the robot samples, when there is the audio data fragment than environmental noise threshold values more peak value, it segment will just be sampled is sent into sound identification module and carry out formal speech recognition, until the peak value of newest audio data is lower than noise threshold values, is considered as segment and terminates.The invention has the advantages that can adapt to environmental change, environmental noise is differentiated in real time, is solved influence of the invalid audio data to sound identification module, is improved machine human efficiency.
Description
Technical field
The present invention relates to a kind of robot more particularly to a kind of machines that can be differentiated environmental noise in real time and carry out speech recognition
People.
Background technology
The sound identification module and order word identification module of existing robot are difficult to correctly identify under noisy environment.Though
Right above-mentioned module has carried the processing such as noise reduction, extraction phonetic feature in itself, but once enter identification state, and module is in itself
Real-time will decline to a great extent or even identification be caused to malfunction because meaningless noise is excessive.And audio is carried out in itself
The method of identifying system excessively redundancy again is re-fed into after noise reduction, the extraction processing such as phonetic feature, computer performance is required it is very high,
And also there is audio distortion after noise reduction in this method.
Invention content
To overcome the above disadvantages, the present invention provides a kind of machines that can be differentiated environmental noise in real time and carry out speech recognition
Device people.
Present invention technical solution used for the above purpose is：
A kind of robot that can be differentiated environmental noise in real time and carry out speech recognition, it is characterized in that：
The robot is analyzed by real-time sampling and the peak value of volume is represented in audio data, by for a long time, regularity
Peak value is included by environmental noise threshold values；The robot, which is analyzed by real-time sampling in audio data, there are new long-time, rule
Existing for rule property during peak value, new peak value is updated to new environmental noise threshold values.Find occur when the robot samples
Than environmental noise threshold values more peak value audio data fragment when, will just sample segment be sent into sound identification module carry out it is formal
Speech recognition until the peak value of newest audio data is lower than noise threshold values, is considered as segment and terminates.
The invention has the advantages that can adapt to environmental change, environmental noise is differentiated in real time, solves invalid audio
Influence of the data to sound identification module, improves machine human efficiency.
Description of the drawings
The present invention is further described with implementation below in conjunction with the accompanying drawings.Fig. 1 is the block diagram of the present invention.
In Fig. 1,1 is sampled audio data, and 2 be to judge to determine environmental noise threshold values, and 3 be determining environmental noise valve again
Value,4 be compared with threshold values, and 5 be to sample segment to be sent into the formal speech recognition of sound identification module progress, and 6 be environmental noise valve
Value.
Specific embodiment
In Fig. 1, the robot analyzes the peak value that volume is represented in audio data by real-time sampling audio data 1,
For example peak value existing for 2 second time, regularity is 35, is just determined as environmental noise threshold values 2 35；The robot passes through real-time
Occur in sampling analysis audio data the 2 second new time, regularity existing for peak value 45 when, new peak value 45 is updated to new
Environmental noise threshold values.It is found when the robot samples, during than the more audio data fragment of peak value 55 of environmental noise threshold values 45,
This sampling segment is just sent into sound identification module and carries out formal speech recognition 5, until the peak value 42 of newest audio data compares
Noise threshold values 45 is low, is considered as segment and terminates.
The above described is only a preferred embodiment of the present invention, not making limitation in any form to the present invention, appoint
What without departing from the present invention program content, any simple modification that technical spirit according to the present invention makees above example, etc.
With variation and modification, in the range of still falling within technical solution of the present invention.
Claims (2)
1. the robot that environmental noise carries out speech recognition can be differentiated in real time, it is characterized in that：
The robot analyzes the peak value that volume is represented in audio data by real-time sampling, and long-time, regularity are deposited
Peak value be included by environmental noise threshold values；Find occur than environmental noise threshold values more peak value when the robot samples
Audio data fragment when, will just sample segment and be sent into sound identification module and carry out formal speech recognition, until newest audio
The peak value of data is lower than noise threshold values, is considered as segment and terminates.
2. the robot according to claim 1 that environmental noise can be differentiated in real time and carry out speech recognition, it is characterized in that：It is described
Robot by real-time sampling analyze audio data in occur new long-time, regularity existing for peak value when, by new peak value
It is updated to new environmental noise threshold values.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611103797.3A CN108154880A (en) | 2016-12-05 | 2016-12-05 | The robot that environmental noise carries out speech recognition can be differentiated in real time |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611103797.3A CN108154880A (en) | 2016-12-05 | 2016-12-05 | The robot that environmental noise carries out speech recognition can be differentiated in real time |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108154880A true CN108154880A (en) | 2018-06-12 |
Family
ID=62469920
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611103797.3A Pending CN108154880A (en) | 2016-12-05 | 2016-12-05 | The robot that environmental noise carries out speech recognition can be differentiated in real time |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108154880A (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5983186A (en) * | 1995-08-21 | 1999-11-09 | Seiko Epson Corporation | Voice-activated interactive speech recognition device and method |
| CN101821971A (en) * | 2007-08-22 | 2010-09-01 | 杜比实验室特许公司 | Systems and methods for noisy activity detection |
| CN101859583A (en) * | 2009-04-13 | 2010-10-13 | 索尼公司 | Noise reducing apparatus and noise reducing method |
| CN102541509A (en) * | 2012-01-13 | 2012-07-04 | 河南科技大学 | True random number generation method based on chaotic encryption |
| CN103117067A (en) * | 2013-01-19 | 2013-05-22 | 渤海大学 | Voice endpoint detection method under low signal-to-noise ratio |
| CN104519447A (en) * | 2013-10-08 | 2015-04-15 | 三星电子株式会社 | Apparatus and method of reducing noise and audio playing apparatus with non-magnet speaker |
| CN104751845A (en) * | 2015-03-31 | 2015-07-01 | 江苏久祥汽车电器集团有限公司 | Voice recognition method and system used for intelligent robot |
| CN104942810A (en) * | 2015-04-29 | 2015-09-30 | 成都陌云科技有限公司 | Intelligent robot capable of denoising |
| CN105611461A (en) * | 2016-01-04 | 2016-05-25 | 浙江宇视科技有限公司 | Noise suppression method, apparatus and system for voice application system of front-end device |
| CN105850154A (en) * | 2013-12-20 | 2016-08-10 | 微软技术许可有限责任公司 | Adapting audio based upon detected environmental acoustics |
| CN105845135A (en) * | 2015-01-12 | 2016-08-10 | 芋头科技（杭州）有限公司 | Sound recognition system and method for robot system |
-
2016
- 2016-12-05 CN CN201611103797.3A patent/CN108154880A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5983186A (en) * | 1995-08-21 | 1999-11-09 | Seiko Epson Corporation | Voice-activated interactive speech recognition device and method |
| CN101821971A (en) * | 2007-08-22 | 2010-09-01 | 杜比实验室特许公司 | Systems and methods for noisy activity detection |
| CN101859583A (en) * | 2009-04-13 | 2010-10-13 | 索尼公司 | Noise reducing apparatus and noise reducing method |
| CN102541509A (en) * | 2012-01-13 | 2012-07-04 | 河南科技大学 | True random number generation method based on chaotic encryption |
| CN103117067A (en) * | 2013-01-19 | 2013-05-22 | 渤海大学 | Voice endpoint detection method under low signal-to-noise ratio |
| CN104519447A (en) * | 2013-10-08 | 2015-04-15 | 三星电子株式会社 | Apparatus and method of reducing noise and audio playing apparatus with non-magnet speaker |
| CN105850154A (en) * | 2013-12-20 | 2016-08-10 | 微软技术许可有限责任公司 | Adapting audio based upon detected environmental acoustics |
| CN105845135A (en) * | 2015-01-12 | 2016-08-10 | 芋头科技（杭州）有限公司 | Sound recognition system and method for robot system |
| CN104751845A (en) * | 2015-03-31 | 2015-07-01 | 江苏久祥汽车电器集团有限公司 | Voice recognition method and system used for intelligent robot |
| CN104942810A (en) * | 2015-04-29 | 2015-09-30 | 成都陌云科技有限公司 | Intelligent robot capable of denoising |
| CN105611461A (en) * | 2016-01-04 | 2016-05-25 | 浙江宇视科技有限公司 | Noise suppression method, apparatus and system for voice application system of front-end device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021128741A1 (en) | Voice emotion fluctuation analysis method and apparatus, and computer device and storage medium | |
| CN108962227B (en) | Voice starting point and end point detection method and device, computer equipment and storage medium | |
| WO2018145584A1 (en) | Voice activity detection method and voice recognition method | |
| JP6229046B2 (en) | Speech data recognition method, device and server for distinguishing local rounds | |
| CN108022587B (en) | Speech recognition method, speech recognition device, computer equipment and storage medium | |
| CN105913849B (en) | A kind of speaker's dividing method based on event detection | |
| JP2014510942A5 (en) | ||
| CN110060693A (en) | Model training method and device, electronic equipment and storage medium | |
| GB201005913D0 (en) | Method for real-time data compression and transmission | |
| WO2006007290B1 (en) | Method and apparatus for equalizing a speech signal generated within a self-contained breathing apparatus system | |
| CN103400576B (en) | Based on speech model update method and the device of User action log | |
| CN106098079B (en) | Method and device for extracting audio signal | |
| CN106448659B (en) | A kind of sound end detecting method based on short-time energy and fractal dimension | |
| WO2018095167A1 (en) | Voiceprint identification method and voiceprint identification system | |
| WO2021127990A1 (en) | Voiceprint recognition method based on voice noise reduction and related apparatus | |
| CN107132311A (en) | A kind of fast gas recognizer extracted based on frequency domain character | |
| CN110689885A (en) | Machine-synthesized speech recognition method, device, storage medium and electronic equipment | |
| Sharma et al. | Automatic identification of silence, unvoiced and voiced chunks in speech | |
| CN107767860A (en) | A kind of voice information processing method and device | |
| CN108154880A (en) | The robot that environmental noise carries out speech recognition can be differentiated in real time | |
| CN107121530B (en) | The electronic nose response spectra feature extracting method that a kind of part and global characteristics combine | |
| CN117935862A (en) | A system for extracting key information from speech | |
| TW200811833A (en) | Detection method for voice activity endpoint | |
| CN103438983B (en) | Data processing method of signal random average spectrum | |
| CN107993666B (en) | Speech recognition method, speech recognition device, computer equipment and readable storage medium |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180612 |