CN110364147B - Awakening training word acquisition system and method - Google Patents
Awakening training word acquisition system and method Download PDFInfo
- Publication number
- CN110364147B CN110364147B CN201910810723.0A CN201910810723A CN110364147B CN 110364147 B CN110364147 B CN 110364147B CN 201910810723 A CN201910810723 A CN 201910810723A CN 110364147 B CN110364147 B CN 110364147B
- Authority
- CN
- China
- Prior art keywords
- awakening
- data terminal
- training
- mobile data
- voice
- 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.)
- Active
Links
- 238000012549 training Methods 0.000 title claims abstract description 115
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000004458 analytical method Methods 0.000 claims abstract description 18
- 230000007613 environmental effect Effects 0.000 claims description 12
- 238000005070 sampling Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000012544 monitoring process Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 4
- 238000005516 engineering process Methods 0.000 description 4
- 241001672694 Citrus reticulata Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000009545 invasion Effects 0.000 description 1
- 230000002618 waking effect Effects 0.000 description 1
Images
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
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
-
- 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
- G10L15/26—Speech to text systems
-
- 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
- G10L15/28—Constructional details of speech recognition systems
- G10L15/30—Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
-
- 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
- G10L15/28—Constructional details of speech recognition systems
- G10L15/34—Adaptation of a single recogniser for parallel processing, e.g. by use of multiple processors or cloud computing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Mathematical Physics (AREA)
- Theoretical Computer Science (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Electrically Operated Instructional Devices (AREA)
- Telephone Function (AREA)
Abstract
The invention belongs to the technical field of voice exchange equipment and particularly relates to a system and a method for acquiring awakening training words, and the system and the method specifically comprise a mobile data terminal, a cloud storage device and a voice recognition cloud platform; the mobile data terminal carries out preliminary judgment on the collected voice data of the specific awakening training words; if the initial judgment is valid, the mobile data terminal submits the collected voice data of the specific awakening training words to a voice recognition cloud platform for analysis to obtain corresponding texts; the text returns to the mobile data terminal; the mobile data terminal matches the text with the voice data of the specific awakening training words; and if the specific awakening word sound data is judged to be valid, the mobile data terminal takes the specific awakening word sound data as a sample awakening word. Has the advantages that: through open collection mode, overcome and to have the inconvenient and speech quality unstability scheduling problem of collection among the current training word collection process of awakening up, constantly strengthen awakening up the word training effect, promote the system to awakening up word recognition accuracy.
Description
Technical Field
The invention belongs to the field of voice exchange equipment and technology, and particularly relates to a system and a method for acquiring awakening training words.
Background
With the development of artificial intelligence and big data technology, the speech recognition technology is more and more widely applied in the fields of vehicle-mounted systems, robots, home services, banks, medical treatment, industrial control and the like by virtue of the advantages of low cost and algorithm complexity, strong calculation power and the like.
The voice awakening word is a voice password for triggering the awakening function of the interactive equipment, and the voice awakening refers to a process that a user realizes that the equipment is recovered from a standby state to a normal working state by speaking a preset awakening word. The accurate voice wake-up word sent by the user is one of key links in the interaction process.
The collection of the awakening training words in the voice awakening technology is particularly important for the training effect, and the accuracy of the system for identifying the awakening words is directly determined. However, the prior art has at least the following defects:
firstly, the identification precision of the mandarin with dialect and dialect accent is not high;
secondly, because of numerous dialects in China, part of dialects with few users can not be identified at present;
thirdly, if the corpus collection is only dependent on the laboratory collection progress, the corpus collection is very slow and the vocabulary amount is limited; if the open authority permits the collection for the respective operation voice platform, although the data can be widely collected, the invasion of the user privacy can be involved;
and fourthly, the corpora are obtained by butting with the existing self-supporting voice platform, wherein the accuracy of free resources is limited, and the paid resources have high cost although the effect is good.
Disclosure of Invention
In order to solve the technical problems, the invention provides a system and a method for acquiring the awakening training words, which overcome the problems of inconvenient acquisition of the awakening training words, unstable acquired audio quality and the like by optimizing the system structure and the method steps, improve the convenience of the process for acquiring the sound data of the awakening training words through open acquisition, meet the personalized requirements of users and facilitate the acquisition of audio data with deep value.
The specific technical scheme is as follows: a wake-up training word acquisition system comprises a mobile data terminal, a cloud storage device and a voice recognition cloud platform;
the mobile data terminal carries out preliminary judgment on the collected voice data of the specific awakening training words;
if the initial judgment is valid, the mobile data terminal submits the collected voice data of the specific awakening training words to a voice recognition cloud platform for analysis to obtain corresponding texts; the text returns to the mobile data terminal;
the mobile data terminal matches the text with the voice data of the specific awakening training words and judges whether the voice data of the specific awakening training words is valid or not according to the matching degree;
and if the sound data of the specific awakening training words is judged to be valid, the mobile data terminal sends the sound data of the specific awakening training words to the cloud storage device to be kept as sample awakening words.
Preferably, the awakening training word acquisition method based on the awakening training word acquisition system includes the following steps:
step one (S101), a user starts 1 or more than 1 mobile data terminal and initializes system parameters; the system parameters comprise a recording authority and a geographic position authority;
step two (S102), 1 or more than 1 mobile data terminal downloads and wakes up training word linguistic data from the cloud storage equipment;
step three (S103), before collecting the voice data of the awakening training words, the mobile data terminal collects the environmental noise voice frequency in advance and submits the environmental noise voice frequency to the cloud storage device;
step four (S104), 1 or more than 1 mobile data terminal collects the voice data of the awakening training words;
step five (S105), the mobile data terminal conducts preliminary analysis processing on the collected awakening training word sound data, and whether the collected awakening training word sound data can be used as a sample awakening word or not is preliminarily judged; if the initial judgment can be used as a sample awakening word, the mobile data terminal submits the collected voice data of the specific awakening training word to a voice recognition cloud platform for analysis to obtain a corresponding text; the text returns to the mobile data terminal; the mobile data terminal matches the text with the voice data of the specific awakening training words and judges the text to be used as sample awakening words according to the matching degree;
step six (S106), if the voice data can be used as the sample awakening words, the mobile data terminal sends the voice data of the awakening training words to the cloud storage device to be kept as the sample awakening words;
step seven (S107), the mobile data terminal repeatedly collects the sample awakening words according to preset times and submits the sample awakening words to the cloud storage device;
and step eight (S108), the mobile data terminal finishes the collection of the voice data of the awakening training words.
Preferably, in the system and method for acquiring the wake-up training words, the first step specifically is: the recording authority is used for starting a recording function of the mobile data terminal to acquire voice recording data; the geographic position authority is used for starting a GPS functional module of the mobile data terminal to record the geographic position information of the voice recording data.
Preferably, in the system and method for acquiring the wake-up training words, the second step is specifically: and the mobile data terminal downloads the awakening training word corpus, corpus codes and corpus recording times from the cloud storage device.
Preferably, in the system and method for acquiring the wake-up training words, the third step is specifically: before the training word sound data is collected every time, the mobile data terminal prompts a user not to sound within 10 seconds, and collects a period of 10 seconds of environmental noise; after the acquisition is finished, storing the file and uploading the environmental noise recording data to the cloud storage device;
noise filename definition:
the corpus code + "_" + mobile data terminal DeviceID + "_" + NOISE + "_" + timestamp.
Preferably, in the system and method for acquiring the wake-up training words, the fourth step is specifically: after monitoring a command for collecting and awakening the training word voice data, the mobile data terminal initializes the recording parameters; the recording parameters comprise audio sampling frequency, sound channel setting, coding system and sampling size; after a user inputs the linguistic data, the mobile data terminal stops audio acquisition in time according to VAD calculation, then intercepts effective audio according to sound energy, and then saves files;
awakening training word voice data filename definition:
the corpus code + "_" + mobile data terminal DeviceID + "_" + round number + "_" + timestamp.
Preferably, in the system and method for acquiring the wake-up training words, the fifth step is specifically: the mobile data terminal judges the effectiveness of the voice data of the awakening training words, namely, invalid voice data is eliminated; if the preliminary analysis is valid, the mobile data terminal submits the collected voice data of the awakening training words to a voice recognition cloud platform for analysis; the voice recognition cloud platform analyzes the voice data to obtain a corresponding text, and returns an analysis result of the mobile data terminal; and the mobile data terminal matches the returned text with the corpus, if the similarity exceeds 70 percent, the mobile data terminal is considered to be effective, and if the similarity does not exceed 70 percent, the mobile data terminal is not effective.
Preferably, in the system and method for acquiring the wake-up training words, the sixth step is specifically: and the mobile data terminal submits the voice data of the specific awakening training words to the cloud storage device to be stored as a sample under the condition that the voice data of the specific awakening training words is judged to be effective.
Preferably, in the system and method for acquiring the wake-up training words, the seventh step is specifically: and the mobile data terminal collects the voice data of the awakening training words according to the preset times, namely, the step four (S104) to the step six (S106) are repeatedly executed.
Preferably, in the system and method for acquiring the wake-up training words, the eighth step specifically is: and the user mobile data terminal finishes the sound data acquisition of the specific awakening training words, and restarts the sound data acquisition of other specific awakening training words from the step four (S104).
Preferably, in the system and the method for acquiring the awakening training words, the mobile data terminal is a smart phone, a tablet computer, a notebook computer or intelligent wearable equipment.
Preferably, in the system and the method for acquiring the awakening training words, the user controls the smart phone by controlling the APP.
After the technical scheme is adopted, the invention has the beneficial effects that:
compared with the disclosed technical scheme, the technical scheme provides the awakening training word acquisition system and method, the system and method steps are optimized, the problems that the existing awakening training word acquisition process is inconvenient to acquire and unstable in acquired voice quality are solved through an open acquisition mode, the awakening words determined as the sample awakening words are acquired for multiple times, the awakening word training effect is continuously enhanced, the identification precision of the system on the awakening words is improved, the personalized requirements of users can be met, and the audio data with deep values can be acquired.
Drawings
FIG. 1 is a schematic diagram of a wake-up training word acquisition system;
FIG. 2 is a flowchart illustrating the steps of a method for waking up a training word acquisition.
Detailed Description
The system shown in fig. 1 includes a mobile data terminal, a cloud storage device, and a voice recognition cloud platform;
the mobile data terminal carries out preliminary judgment on the collected voice data of the specific awakening training words;
if the initial judgment is valid, the mobile data terminal submits the collected voice data of the specific awakening training words to a voice recognition cloud platform for analysis to obtain corresponding texts; the text returns to the mobile data terminal;
the mobile data terminal matches the text with the voice data of the specific awakening training words and judges whether the voice data of the specific awakening training words is valid or not according to the matching degree;
and if the sound data of the specific awakening training words is judged to be valid, the mobile data terminal sends the sound data of the specific awakening training words to the cloud storage device to be kept as sample awakening words.
The mobile data terminal can be a smart phone, a tablet computer, a notebook computer, an intelligent wearable device and the like; the mobile data terminal of the embodiment is exemplified by a smart phone.
Referring to fig. 2, a method for collecting a wake-up training word mainly includes the following steps:
step one (S101), a user starts 1 or more than 1 smart phone APP and initializes system parameters; the system parameters comprise a recording authority and a geographic position authority; the recording authority is used for starting a recording function of the smart phone to acquire voice recording data; the geographic position authority is used for starting a GPS functional module of the smart phone to record geographic position information of the voice recording data.
And step two (S102), 1 or more than 1 smart phone downloads and wakes up the corpus of the training words, corpus codes and corpus recording times from the cloud storage device.
Step three (S103), before collecting the voice data of the awakening training words, the smart phone collects the environmental noise voice frequency in advance and submits the environmental noise voice frequency to the cloud storage device; before the training word sound data is collected every time, the smart phone prompts a user not to sound within 10 seconds, and collects a period of 10 seconds of environmental noise; after the acquisition is finished, storing the file and uploading the environmental noise recording data to the cloud storage device;
noise filename definition:
the corpus code + "_" + smartphone DeviceID + "_" + NOISE + "_" + timestamp.
Step four (S104), 1 or more than 1 smart phone collects and awakens training word sound data; after monitoring a command for collecting and awakening training word voice data, the smart phone initializes recording parameters; the recording parameters comprise audio sampling frequency, sound channel setting, coding system and sampling size; after a user inputs the linguistic data, the smart phone stops audio acquisition in time according to VAD calculation, then intercepts effective audio according to sound energy, and then saves files;
awakening training word voice data filename definition:
the corpus code + "_" + smartphone DeviceID + "_" + round number + "_" + timestamp.
Step five (S105), the smart phone APP performs primary analysis processing on the collected awakening training word sound data, and the smart phone APP judges the effectiveness of the awakening training word sound data, namely, invalid voice data is eliminated; if the preliminary analysis is valid, the smart phone submits the collected voice data of the awakening training words to a voice recognition cloud platform for analysis; the voice recognition cloud platform analyzes the voice data to obtain a corresponding text, and returns an analysis result of the smart phone; and matching the returned text with the corpus by the smart phone, and if the similarity exceeds 70%, determining that the returned text is valid, otherwise, determining that the returned text is invalid.
And step six (S106), under the condition that the specific awakening training word sound data are judged to be effective, the smart phone APP submits the specific awakening training word sound data to the cloud storage device to be stored as a sample.
Step seven (S107), repeatedly collecting the sample awakening words by the smart phone APP according to preset times, and submitting the sample awakening words to cloud storage equipment; and the smart phone APP collects the voice data of the awakening training words according to the preset times, namely, the step four (S104) to the step six (S106) are repeatedly executed.
And step eight (S108), the smart phone finishes awakening training word sound data acquisition. And the user smart phone finishes the sound data collection of the specific awakening training words, and restarts the collection of the sound data of other specific awakening training words from the step four (S104).
It will be appreciated by those skilled in the art that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The embodiments disclosed above are therefore to be considered in all respects as illustrative and not restrictive. All changes which come within the scope of or equivalence to the invention are intended to be embraced therein.
Claims (7)
1. A wake-up training word acquisition system comprises a mobile data terminal, a cloud storage device and a voice recognition cloud platform; the method is characterized in that:
the mobile data terminal carries out preliminary judgment on the collected voice data of the specific awakening training words;
if the initial judgment is valid, the mobile data terminal submits the collected voice data of the specific awakening training words to a voice recognition cloud platform for analysis to obtain corresponding texts; the text returns to the mobile data terminal;
the mobile data terminal matches the text with the voice data of the specific awakening training words and judges whether the voice data of the specific awakening training words is valid or not according to the matching degree;
if the voice data of the specific awakening training words are judged to be valid, the mobile data terminal sends the voice data of the specific awakening training words to the cloud storage device to be kept as sample awakening words;
the awakening training word acquisition method based on the awakening training word acquisition system comprises the following steps of:
step one (S101), a user starts 1 or more than 1 mobile data terminal and initializes system parameters; the system parameters comprise a recording authority and a geographic position authority;
step two (S102), 1 or more than 1 mobile data terminal downloads and wakes up training word linguistic data from the cloud storage equipment;
step three (S103), before collecting the voice data of the awakening training words, the mobile data terminal collects the environmental noise voice frequency in advance and submits the environmental noise voice frequency to the cloud storage device; the method specifically comprises the following steps: before the training word sound data is collected every time, the mobile data terminal prompts a user not to sound within 10 seconds, and collects a period of 10 seconds of environmental noise; after the acquisition is finished, storing the file and uploading the environmental noise recording data to the cloud storage device;
noise filename definition:
a corpus code + "_" + mobile data terminal DeviceID + "_" + NOISE + "_" + timestamp;
step four (S104), 1 or more than 1 mobile data terminal collects the voice data of the awakening training words;
step five (S105), the mobile data terminal conducts preliminary analysis processing on the collected awakening training word sound data, and whether the collected awakening training word sound data can be used as a sample awakening word or not is preliminarily judged; if the initial judgment can be used as a sample awakening word, the mobile data terminal submits the collected voice data of the specific awakening training word to a voice recognition cloud platform for analysis to obtain a corresponding text; the text returns to the mobile data terminal; the mobile data terminal matches the text with the voice data of the specific awakening training words and judges the text to be used as sample awakening words according to the matching degree; the method specifically comprises the following steps: the mobile data terminal judges the effectiveness of the voice data of the awakening training words, namely, invalid voice data is eliminated; if the preliminary analysis is valid, the mobile data terminal submits the collected voice data of the awakening training words to a voice recognition cloud platform for analysis; the voice recognition cloud platform analyzes the voice data to obtain a corresponding text, and returns an analysis result of the mobile data terminal; the mobile data terminal matches the returned text with the corpus, if the similarity exceeds 70%, the mobile data terminal is considered to be effective, otherwise, the mobile data terminal is invalid;
step six (S106), if the voice data can be used as the sample awakening words, the mobile data terminal sends the voice data of the awakening training words to the cloud storage device to be kept as the sample awakening words;
step seven (S107), the mobile data terminal repeatedly collects the sample awakening words according to preset times and submits the sample awakening words to the cloud storage device;
and step eight (S108), the mobile data terminal finishes the collection of the voice data of the awakening training words.
2. The system for acquiring wake-up training words according to claim 1, wherein the first step is specifically: the recording authority is used for starting a recording function of the mobile data terminal to acquire voice recording data; and the geographic position authority is used for starting a GPS functional module of the mobile data terminal to record the geographic position information of the voice recording data.
3. The system for acquiring the awakening training words according to claim 1, wherein the second step is specifically as follows: and the mobile data terminal downloads the awakening training word corpus, corpus codes and corpus recording times from the cloud storage device.
4. The system for acquiring the awakening training words according to claim 1, wherein the fourth step is specifically as follows: after monitoring a command for collecting and awakening the training word voice data, the mobile data terminal initializes the recording parameters; the recording parameters comprise audio sampling frequency, sound channel setting, coding system and sampling size; after a user inputs the linguistic data, the mobile data terminal stops audio acquisition in time according to VAD calculation, then intercepts effective audio according to sound energy, and then saves files;
awakening training word voice data filename definition:
the corpus code + "_" + mobile data terminal DeviceID + "_" + round number + "_" + timestamp.
5. The system for acquiring the awakening training words according to claim 1, wherein the sixth step is specifically as follows: and the mobile data terminal submits the voice data of the specific awakening training words to the cloud storage device to be stored as a sample under the condition that the voice data of the specific awakening training words is judged to be effective.
6. The system for acquiring the awakening training words according to claim 1, wherein the seventh step is specifically as follows: and the mobile data terminal collects the voice data of the awakening training words according to the preset times, namely, the step four (S104) to the step six (S106) are repeatedly executed.
7. The system for acquiring the awakening training words according to claim 1, wherein the eighth step is specifically as follows: and the user mobile data terminal finishes the sound data acquisition of the specific awakening training words, and restarts the sound data acquisition of other specific awakening training words from the step four (S104).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910810723.0A CN110364147B (en) | 2019-08-29 | 2019-08-29 | Awakening training word acquisition system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910810723.0A CN110364147B (en) | 2019-08-29 | 2019-08-29 | Awakening training word acquisition system and method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110364147A CN110364147A (en) | 2019-10-22 |
CN110364147B true CN110364147B (en) | 2021-08-20 |
Family
ID=68225569
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910810723.0A Active CN110364147B (en) | 2019-08-29 | 2019-08-29 | Awakening training word acquisition system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110364147B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110970016B (en) * | 2019-10-28 | 2022-08-19 | 苏宁云计算有限公司 | Awakening model generation method, intelligent terminal awakening method and device |
CN111128150A (en) * | 2019-11-27 | 2020-05-08 | 云知声智能科技股份有限公司 | Method and device for awakening intelligent voice equipment |
CN113593549A (en) * | 2021-06-29 | 2021-11-02 | 青岛海尔科技有限公司 | Method and device for determining awakening rate of voice equipment |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103646646A (en) * | 2013-11-27 | 2014-03-19 | 联想(北京)有限公司 | Voice control method and electronic device |
CN106847283A (en) * | 2017-02-28 | 2017-06-13 | 广东美的制冷设备有限公司 | Intelligent electrical appliance control and device |
CN107134279A (en) * | 2017-06-30 | 2017-09-05 | 百度在线网络技术(北京)有限公司 | A kind of voice awakening method, device, terminal and storage medium |
CN107358954A (en) * | 2017-08-29 | 2017-11-17 | 成都启英泰伦科技有限公司 | It is a kind of to change the device and method for waking up word in real time |
CN107704275A (en) * | 2017-09-04 | 2018-02-16 | 百度在线网络技术(北京)有限公司 | Smart machine awakening method, device, server and smart machine |
CN108958810A (en) * | 2018-02-09 | 2018-12-07 | 北京猎户星空科技有限公司 | A kind of user identification method based on vocal print, device and equipment |
CN109036393A (en) * | 2018-06-19 | 2018-12-18 | 广东美的厨房电器制造有限公司 | Wake-up word training method, device and the household appliance of household appliance |
CN109360552A (en) * | 2018-11-19 | 2019-02-19 | 广东小天才科技有限公司 | Method and system for automatically filtering awakening words |
CN109584860A (en) * | 2017-09-27 | 2019-04-05 | 九阳股份有限公司 | A kind of voice wakes up word and defines method and system |
CN109949808A (en) * | 2019-03-15 | 2019-06-28 | 上海华镇电子科技有限公司 | The speech recognition appliance control system and method for compatible mandarin and dialect |
CN110097876A (en) * | 2018-01-30 | 2019-08-06 | 阿里巴巴集团控股有限公司 | Voice wakes up processing method and is waken up equipment |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160267380A1 (en) * | 2015-03-13 | 2016-09-15 | Nuance Communications, Inc. | Method and System for Training a Neural Network |
US10388274B1 (en) * | 2016-03-31 | 2019-08-20 | Amazon Technologies, Inc. | Confidence checking for speech processing and query answering |
US10332517B1 (en) * | 2017-06-02 | 2019-06-25 | Amazon Technologies, Inc. | Privacy mode based on speaker identifier |
CN109817220A (en) * | 2017-11-17 | 2019-05-28 | 阿里巴巴集团控股有限公司 | Audio recognition method, apparatus and system |
-
2019
- 2019-08-29 CN CN201910810723.0A patent/CN110364147B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103646646A (en) * | 2013-11-27 | 2014-03-19 | 联想(北京)有限公司 | Voice control method and electronic device |
CN106847283A (en) * | 2017-02-28 | 2017-06-13 | 广东美的制冷设备有限公司 | Intelligent electrical appliance control and device |
CN107134279A (en) * | 2017-06-30 | 2017-09-05 | 百度在线网络技术(北京)有限公司 | A kind of voice awakening method, device, terminal and storage medium |
CN107358954A (en) * | 2017-08-29 | 2017-11-17 | 成都启英泰伦科技有限公司 | It is a kind of to change the device and method for waking up word in real time |
CN107704275A (en) * | 2017-09-04 | 2018-02-16 | 百度在线网络技术(北京)有限公司 | Smart machine awakening method, device, server and smart machine |
CN109584860A (en) * | 2017-09-27 | 2019-04-05 | 九阳股份有限公司 | A kind of voice wakes up word and defines method and system |
CN110097876A (en) * | 2018-01-30 | 2019-08-06 | 阿里巴巴集团控股有限公司 | Voice wakes up processing method and is waken up equipment |
CN108958810A (en) * | 2018-02-09 | 2018-12-07 | 北京猎户星空科技有限公司 | A kind of user identification method based on vocal print, device and equipment |
CN109036393A (en) * | 2018-06-19 | 2018-12-18 | 广东美的厨房电器制造有限公司 | Wake-up word training method, device and the household appliance of household appliance |
CN109360552A (en) * | 2018-11-19 | 2019-02-19 | 广东小天才科技有限公司 | Method and system for automatically filtering awakening words |
CN109949808A (en) * | 2019-03-15 | 2019-06-28 | 上海华镇电子科技有限公司 | The speech recognition appliance control system and method for compatible mandarin and dialect |
Also Published As
Publication number | Publication date |
---|---|
CN110364147A (en) | 2019-10-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110310623B (en) | Sample generation method, model training method, device, medium, and electronic apparatus | |
CN110838289B (en) | Wake-up word detection method, device, equipment and medium based on artificial intelligence | |
CN110364147B (en) | Awakening training word acquisition system and method | |
CN105632486B (en) | Voice awakening method and device of intelligent hardware | |
EP2940684B1 (en) | Voice recognizing method and system for personalized user information | |
CN110570873B (en) | Voiceprint wake-up method and device, computer equipment and storage medium | |
CN107220532B (en) | Method and apparatus for recognizing user identity through voice | |
CN104575504A (en) | Method for personalized television voice wake-up by voiceprint and voice identification | |
CN108595406B (en) | User state reminding method and device, electronic equipment and storage medium | |
CN113129867B (en) | Training method of voice recognition model, voice recognition method, device and equipment | |
WO2020233363A1 (en) | Speech recognition method and device, electronic apparatus, and storage medium | |
CN111179944B (en) | Voice awakening and age detection method and device and computer readable storage medium | |
CN109074809B (en) | Information processing apparatus, information processing method, and computer-readable storage medium | |
CN114360510A (en) | Voice recognition method and related device | |
CN112669818A (en) | Voice wake-up method and device, readable storage medium and electronic equipment | |
CN115064160B (en) | Voice wake-up method and device | |
CN111179941A (en) | Intelligent device awakening method, registration method and device | |
CN108010518B (en) | Voice acquisition method, system and storage medium of voice interaction equipment | |
Zhang et al. | Research on spectrum sensing system based on composite neural network | |
CN111048068B (en) | Voice wake-up method, device and system and electronic equipment | |
CN101350196A (en) | On-chip system for confirming role related talker identification and confirming method thereof | |
CN115691478A (en) | Voice wake-up method and device, man-machine interaction equipment and storage medium | |
CN112669836B (en) | Command recognition method and device and computer readable storage medium | |
CN114049879A (en) | Speech recognition method based on artificial intelligence algorithm | |
CN113362830A (en) | Starting method, control method, system and storage medium of voice assistant |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |