CN108847219B - Awakening word preset confidence threshold adjusting method and system - Google Patents
Awakening word preset confidence threshold adjusting method and system Download PDFInfo
- Publication number
- CN108847219B CN108847219B CN201810516713.1A CN201810516713A CN108847219B CN 108847219 B CN108847219 B CN 108847219B CN 201810516713 A CN201810516713 A CN 201810516713A CN 108847219 B CN108847219 B CN 108847219B
- Authority
- CN
- China
- Prior art keywords
- user
- time period
- preset confidence
- preset
- threshold
- 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
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000002354 daily effect Effects 0.000 claims description 61
- 230000003203 everyday effect Effects 0.000 claims description 10
- 230000009191 jumping Effects 0.000 claims description 3
- 230000006399 behavior Effects 0.000 description 50
- 230000006870 function Effects 0.000 description 11
- 230000002618 waking effect Effects 0.000 description 4
- 230000007547 defect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 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
- G10L15/22—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
- 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
- G10L15/28—Constructional details of speech recognition 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
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
- G10L2015/0635—Training updating or merging of old and new templates; Mean values; Weighting
- G10L2015/0636—Threshold criteria for the updating
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)
- Artificial Intelligence (AREA)
- Telephone Function (AREA)
- Electric Clocks (AREA)
Abstract
The invention discloses a method for adjusting a preset confidence threshold of an awakening word, which comprises the following steps: s11, acquiring daily use behavior data of a user; and S12, adjusting a preset confidence level threshold value of the awakening words according to the daily use behavior data of the user. The method dynamically sets and adjusts the preset confidence threshold of the awakening words by acquiring the daily use behavior data of the user, fully considers the personal use characteristics of the user, judges whether the preset confidence threshold is integrally adjusted down or not by the daily behavior confidence average value of the user, and further adjusts the preset confidence threshold according to the use times of each time period, thereby effectively improving the equipment awakening rate and reducing the false awakening rate.
Description
Technical Field
The invention relates to the field of voice processing, in particular to a method and a system for adjusting a preset confidence threshold of a wakeup word.
Background
The voice recognition technology has made remarkable progress in recent years, and has entered various fields such as industry, home appliances, smart home, and the like. Voice wakeup is a form of voice recognition technology, which does not directly contact a hardware device, and can wake up the device to run through voice. Voice awakening is to enable devices (smart speakers, mobile phones, toys, household appliances and the like) to detect the voice of a user (a set voice instruction, namely awakening words) in a dormant or screen locking state, enable the devices in the dormant state to directly enter a waiting instruction state, and start a first voice interaction step. In general, most devices rely on physical keys to wake up or operate the devices. However, this is not good for the user experience. Voice, the most natural way of communicating people, waking up a device in a contactless manner by voice is undoubtedly more friendly,
in the prior art, in order to improve the success rate of voice awakening, a voice model is trained, and a proper confidence threshold is set, where the confidence is the matching conformity between the language input of the user and the preset language model, and the higher the confidence is, the more the language input of the user is matched with the preset language model, and the more the language input of the user is mismatched. When the software judges whether the user language behavior wakes up the equipment, the confidence value obtained by the user input is compared with the preset confidence level, if the confidence value is greater than the preset confidence level, the waking is successful, otherwise, the waking fails. However, the confidence threshold value of the voice model training is fixed, the method is only suitable for common user scenes, the threshold values adopted by different users are the same, the personal use characteristics of the different users are not considered, and although the awakening success rate can be improved, the false awakening rate cannot be reduced.
Patent publication No. CN105702253A discloses a voice wake-up method and apparatus for improving accuracy of waking up a terminal device by using voice. The method comprises the following steps: when terminal equipment receives first voice data which are input by a user and contain a preset awakening word, matching the first voice data with a preset language model to obtain a confidence coefficient of the first voice data; judging whether the confidence coefficient is smaller than a preset confidence coefficient threshold value; when the confidence coefficient is smaller than the preset confidence coefficient threshold value, executing a preset operation; and when the confidence coefficient is greater than or equal to the preset confidence coefficient threshold value, awakening the voice control function of the terminal equipment. According to the technical scheme, when the user awakens the terminal device by voice fails, the terminal device can improve the confidence coefficient of the first voice data by executing the preset operation, so that the accuracy of awakening the terminal device by voice and the experience of the user are improved. Although the method improves the awakening success rate, the false awakening rate cannot be reduced, and the confidence threshold value in the method is also fixed, so that the personal use characteristics of different users are not comprehensively considered.
Disclosure of Invention
The invention aims to provide a method and a system for dynamically adjusting a confidence threshold of an awakening word based on user behavior aiming at the defects of the prior art, and solves the problems that the confidence threshold of the prior art is fixed, the personal use characteristics of different users are not considered, and the false awakening rate cannot be reduced.
In order to achieve the purpose, the invention adopts the following technical scheme:
a preset confidence threshold adjusting method for a wakeup word comprises the following steps:
s1, acquiring daily use behavior data of a user;
and S2, adjusting a preset confidence level threshold value of the awakening words according to the daily use behavior data of the user.
Further, the user daily behavior data comprises the confidence value average value of the user daily behavior and the awakening times of the user in each time period every day.
Further, the step S2 is preceded by the steps of:
and determining the daily high-frequency use time period and the daily low-frequency use time period of the user according to the awakening times of the user in each time period.
Further, step S2 is specifically that:
s201, judging whether the user daily behavior confidence coefficient average value is far higher than the preset confidence coefficient threshold value, and if not, jumping to the step S202;
s202, overall adjusting the preset confidence coefficient threshold value.
Further, the step S202 is followed by the steps of:
judging whether the time period of the current time is in the high-frequency use time period or the low-frequency use time period;
and if the preset confidence threshold is in the high-frequency use time period, the preset confidence threshold is adjusted downwards, and if the preset confidence threshold is in the low-frequency use time period, the preset confidence threshold is adjusted upwards.
Correspondingly, a preset confidence threshold adjusting system for awakening words comprises:
the data acquisition module is used for acquiring daily use behavior data of a user;
and the adjusting module is used for adjusting the preset confidence level threshold of the awakening words according to the daily use behavior data of the user.
Further, the user daily behavior data comprises the confidence value average value of the user daily behavior and the awakening times of the user in each time period every day.
Further, the method also comprises the following steps:
and the high-low frequency use time period determining module is used for determining the high-frequency use time period and the low-frequency use time period of each day of the user according to the awakening times of each time period of each day of the user.
Further, the adjusting module comprises:
the first judgment module is used for judging whether the user daily behavior confidence coefficient average value is far higher than the preset confidence coefficient threshold value;
a first adjusting module, configured to integrally adjust the preset confidence threshold downward when the user daily behavior confidence average value is not much higher than the preset confidence threshold.
Further, the adjusting module further comprises:
the second judging module is used for judging whether the time period of the current time is in the high-frequency using time period or the low-frequency using time period;
and the second adjusting module is used for adjusting the preset confidence threshold value downwards when the current time period is in the high-frequency using time period, and adjusting the preset confidence threshold value upwards when the current time period is in the low-frequency using time period.
Compared with the prior art, the method and the device have the advantages that the preset confidence level threshold of the awakening words is dynamically set and adjusted by acquiring the daily use behavior data of the user, the personal use characteristics of the user are fully considered, whether the preset confidence level threshold is integrally adjusted down is judged by the daily behavior confidence level average value of the user, the preset confidence level threshold is further adjusted according to the use times of the user in each time period, and the equipment awakening rate is effectively improved and the mistaken awakening rate is reduced.
Drawings
Fig. 1 is a flowchart of a method for adjusting a preset confidence threshold of a wakeup word according to an embodiment;
fig. 2 is a structural diagram of a system for adjusting a preset confidence threshold of a wake-up word according to the second embodiment.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.
The voice product in daily life is taken as a research object, the main focus is to improve the awakening rate of the voice product, reduce the false awakening rate and optimize the use experience of a user on the voice product. Aiming at the defects of the prior art, the invention provides a method and a system for dynamically adjusting the confidence threshold of the awakening word based on user behavior, solves the problems that the confidence threshold of the prior art is fixed, the personal use characteristics of different users are not considered, the false awakening rate cannot be reduced, and improves the user experience.
Example one
The embodiment provides a method for adjusting a preset confidence threshold of a wakeup word, as shown in fig. 1, including the steps of:
s11, acquiring daily use behavior data of a user;
and S12, adjusting a preset confidence level threshold value of the awakening words according to the daily use behavior data of the user.
The method for adjusting the preset confidence threshold of the wake-up word in the embodiment is applied to a terminal device, and the terminal device may be any device with a voice control function, such as a mobile phone, a computer, an intelligent sound box, a toy, a household appliance, a digital broadcast terminal, a message transceiver, a tablet device, a medical device, a fitness device, and a personal digital assistant.
The main execution body of the method for adjusting the preset confidence level threshold of the wake-up word in the embodiment is the terminal device.
It should be noted that the wake-up word is a word related to the voice control function of the terminal device and is preset by the user. For example, if the voice control function of the terminal device includes controlling an intelligent home, the preset wake-up word may include words related to the intelligent home, such as an air conditioner, a television, a curtain, and the like; for another example, if the voice control function of the terminal device includes connecting to a cloud server and searching for network information through the cloud server, the preset wake-up word may include words related to network services such as search, query, weather, train ticket, and the like.
Specifically, before the user uses the terminal device, the user is required to upload the gender, the age, the time period of using the terminal device, the function preference of the terminal device and the like to obtain the basic use behavior information of the user to set the preset confidence level threshold size of the awakening word. And then acquiring daily use behavior data of the user, and adjusting the preset confidence level of the awakening words according to the daily use behavior data of the user.
Optionally, the user daily behavior data includes a user daily behavior confidence average value and the number of awakenings of the user in each time period per day.
Optionally, before the step S12, the method further includes the steps of:
and determining the daily high-frequency use time period and the daily low-frequency use time period of the user according to the awakening times of the user in each time period.
Specifically, the terminal device may calculate the confidence level average value of the daily behavior of the user at 24 points per day, where the calculation process specifically includes: and counting the confidence value of each awakening word of the user in the whole day, adding the confidence values of all the awakening words in the next day, and dividing the sum by the number of awakenings to obtain the confidence average value of the daily behavior of the user.
On the other hand, the terminal device counts the awakening times of the user in each time period every day to determine the high-frequency using time period and the low-frequency using time period every day of the user.
Alternatively, each time period may be one time period every two hours. For example, a time period is from 0 to 2, a time period is from 2 to 4, and so on, then there are 12 time periods in a day.
Optionally, the low-frequency usage time period is determined if the usage frequency of the user in one time period is lower than twice, and the high-frequency usage time period is determined if the usage frequency of the user in one time period is higher than ten times. For example: the following table shows the user's various time periods per day and the corresponding number of awakenings.
As can be seen from the above table, the number of times of user awakening in the 4 time periods [0, 2], [2, 4], [4, 6], [6, 8] is less than two, and then the 4 time periods are low frequency usage time periods, and the number of times of user awakening in the 3 time periods [16, 18], [18, 20], [20, 22] is more than ten, and then the 3 time periods are high frequency usage time periods.
Optionally, step S12 specifically includes:
s1201, judging whether the user daily behavior confidence coefficient average value is far higher than the preset confidence coefficient threshold value, and if not, jumping to the step S1202;
and S1202, integrally adjusting the preset confidence coefficient threshold value.
Specifically, after 24 points per day, the terminal device calculates the confidence coefficient average value of the daily behavior of the user, and then compares the confidence coefficient average value of the daily behavior of the user with the preset confidence coefficient threshold value, if the confidence coefficient average value of the daily behavior of the user is far higher than the preset confidence coefficient threshold value, the preset confidence coefficient threshold value is not adjusted, and because the execution degree average value of the daily behavior of the user is far higher than the preset confidence coefficient threshold value, the confidence coefficient value of the voice data input by the user easily reaches the preset confidence coefficient threshold value, and therefore the preset confidence coefficient threshold value does not need to be adjusted. If the average value of the confidence degrees of the daily behaviors of the user is not far higher (close) than the preset confidence degree threshold, the preset confidence degree threshold is integrally adjusted downwards, and the confidence degree value of the voice data input by the user can more easily reach the preset confidence degree threshold through integrally adjusting the preset confidence degree threshold downwards, so that the voice data of the user can be more easily identified, and the awakening rate of the user to the terminal equipment is improved.
Optionally, after the step S1202, the method further includes the step of:
judging whether the time period of the current time is in the high-frequency use time period or the low-frequency use time period;
and if the preset confidence threshold is in the high-frequency use time period, the preset confidence threshold is adjusted downwards, and if the preset confidence threshold is in the low-frequency use time period, the preset confidence threshold is adjusted upwards.
Specifically, the terminal device may further adjust the preset confidence level threshold according to each time period every day, if the current time is a high-frequency usage time period (i.e. a time period in which the number of awakenings exceeds ten), the preset confidence threshold is adjusted downward so that the confidence value of the voice data input by the user more easily reaches the preset confidence threshold, therefore, the awakening rate of the terminal equipment by the user can be improved in the high-frequency use time period, if the current time is in the low-frequency use time period (namely the time period with the awakening times lower than two), because the low-frequency use time period is less in use, the preset confidence threshold value is adjusted upwards, so that the confidence value of the voice data input by the user is not easy to reach the preset confidence threshold value relative to other time, and the false awakening rate of the user to the terminal equipment can be reduced in the low-frequency use time period.
According to the embodiment, the awakening word preset confidence level threshold is dynamically set and adjusted by acquiring daily use behavior data of the user, personal use characteristics of the user are fully considered, whether the preset confidence level threshold is integrally adjusted down is judged by the daily behavior confidence level average value of the user, the preset confidence level threshold is further adjusted according to the use times of each time period, and the awakening rate is effectively improved and the false awakening rate is reduced.
Example two
The present embodiment provides a system for adjusting a preset confidence threshold of a wake-up word, as shown in fig. 2, including:
the data acquisition module 11 is used for acquiring daily use behavior data of a user;
and the adjusting module 12 is configured to adjust a preset confidence threshold of the wake-up word according to the daily usage behavior data of the user.
The method for adjusting the preset confidence threshold of the wake-up word in the embodiment is applied to a terminal device, and the terminal device may be any device with a voice control function, such as a mobile phone, a computer, an intelligent sound box, a toy, a household appliance, a digital broadcast terminal, a message transceiver, a tablet device, a medical device, a fitness device, and a personal digital assistant.
It should be noted that the wake-up word is a word related to the voice control function of the terminal device and is preset by the user. For example, if the voice control function of the terminal device includes controlling an intelligent home, the preset wake-up word may include words related to the intelligent home, such as an air conditioner, a television, a curtain, and the like; for another example, if the voice control function of the terminal device includes connecting to a cloud server and searching for network information through the cloud server, the preset wake-up word may include words related to network services such as search, query, weather, train ticket, and the like.
Specifically, before the user uses the terminal device, the user is required to upload the gender, the age, the time period of using the terminal device, the function preference of the terminal device and the like to obtain the basic use behavior information of the user to set the preset confidence level threshold size of the awakening word. And then acquiring daily use behavior data of the user, and adjusting the preset confidence level of the awakening words according to the daily use behavior data of the user.
Optionally, the user daily behavior data includes a user daily behavior confidence average value and the number of awakenings of the user in each time period per day.
Optionally, the method further includes:
and the high-low frequency use time period determining module is used for determining the high-frequency use time period and the low-frequency use time period of each day of the user according to the awakening times of each time period of each day of the user.
Specifically, the terminal device may calculate the confidence level average value of the daily behavior of the user at 24 points per day, where the calculation process specifically includes: and counting the confidence value of each awakening word of the user in the whole day, adding the confidence values of all the awakening words in the next day, and dividing the sum by the number of awakenings to obtain the confidence average value of the daily behavior of the user.
On the other hand, the terminal device counts the awakening times of the user in each time period every day to determine the high-frequency using time period and the low-frequency using time period every day of the user.
Alternatively, each time period may be one time period every two hours. For example, a time period is from 0 to 2, a time period is from 2 to 4, and so on, then there are 12 time periods in a day.
Optionally, the low-frequency usage time period is determined if the usage frequency of the user in one time period is lower than twice, and the high-frequency usage time period is determined if the usage frequency of the user in one time period is higher than ten times. For example: the following table shows the user's various time periods per day and the corresponding number of awakenings.
As can be seen from the above table, the number of times of user awakening in the 4 time periods [0, 2], [2, 4], [4, 6], [6, 8] is less than two, and then the 4 time periods are low frequency usage time periods, and the number of times of user awakening in the 3 time periods [16, 18], [18, 20], [20, 22] is more than ten, and then the 3 time periods are high frequency usage time periods.
Optionally, the adjusting module includes:
the first judgment module is used for judging whether the user daily behavior confidence coefficient average value is far higher than the preset confidence coefficient threshold value;
a first adjusting module, configured to integrally adjust the preset confidence threshold downward when the user daily behavior confidence average value is not much higher than the preset confidence threshold.
Specifically, after 24 points per day, the terminal device calculates the confidence coefficient average value of the daily behavior of the user, and then compares the confidence coefficient average value of the daily behavior of the user with the preset confidence coefficient threshold value, if the confidence coefficient average value of the daily behavior of the user is far higher than the preset confidence coefficient threshold value, the preset confidence coefficient threshold value is not adjusted, and because the execution degree average value of the daily behavior of the user is far higher than the preset confidence coefficient threshold value, the confidence coefficient value of the voice data input by the user easily reaches the preset confidence coefficient threshold value, and therefore the preset confidence coefficient threshold value does not need to be adjusted. If the average value of the confidence degrees of the daily behaviors of the user is not far higher (close) than the preset confidence degree threshold, the preset confidence degree threshold is integrally adjusted downwards, and the confidence degree value of the voice data input by the user can more easily reach the preset confidence degree threshold through integrally adjusting the preset confidence degree threshold downwards, so that the voice data of the user can be more easily identified, and the awakening rate of the user to the terminal equipment is improved.
Optionally, the adjusting module further includes:
the second judging module is used for judging whether the time period of the current time is in the high-frequency using time period or the low-frequency using time period;
and the second adjusting module is used for adjusting the preset confidence threshold value downwards when the current time period is in the high-frequency using time period, and adjusting the preset confidence threshold value upwards when the current time period is in the low-frequency using time period.
Specifically, the terminal device may further adjust the preset confidence level threshold according to each time period every day, if the current time is a high-frequency usage time period (i.e. a time period in which the number of awakenings exceeds ten), the preset confidence threshold is adjusted downward so that the confidence value of the voice data input by the user more easily reaches the preset confidence threshold, therefore, the awakening rate of the terminal equipment by the user can be improved in the high-frequency use time period, if the current time is in the low-frequency use time period (namely the time period with the awakening times lower than two), because the low-frequency use time period is less in use, the preset confidence threshold value is adjusted upwards, so that the confidence value of the voice data input by the user is not easy to reach the preset confidence threshold value relative to other time, and the false awakening rate of the user to the terminal equipment can be reduced in the low-frequency use time period.
According to the embodiment, the data acquisition module is used for acquiring the daily use behavior data of the user to dynamically set and adjust the preset confidence level threshold of the awakening words, the personal use characteristics of the user are fully considered, the awakening rate is effectively improved, the mistaken awakening rate is reduced, and the user experience is improved.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (2)
1. A preset confidence threshold adjusting method for a wakeup word is characterized by comprising the following steps:
s1, acquiring daily use behavior data of a user;
s2, adjusting a preset confidence level threshold value of the awakening words according to the daily use behavior data of the user;
the user daily behavior data comprise a user daily behavior confidence coefficient average value and the awakening times of the user in each time period every day;
the step S2 is preceded by the steps of:
determining a high-frequency use time period and a low-frequency use time period of each day of the user according to the awakening times of each time period of each day of the user;
the step S2 specifically includes:
s201, judging whether the user daily behavior confidence coefficient average value is far higher than the preset confidence coefficient threshold value, and if not, jumping to the step S202;
s202, integrally adjusting the preset confidence level threshold value;
the step S202 is followed by the step of:
judging whether the time period of the current time is in the high-frequency use time period or the low-frequency use time period;
and if the preset confidence threshold is in the high-frequency use time period, the preset confidence threshold is adjusted downwards, and if the preset confidence threshold is in the low-frequency use time period, the preset confidence threshold is adjusted upwards.
2. A wake word preset confidence threshold adjustment system, comprising:
the data acquisition module is used for acquiring daily use behavior data of a user;
the adjusting module is used for adjusting a preset reliability threshold of the awakening words according to the daily use behavior data of the user;
the user daily behavior data comprise a user daily behavior confidence coefficient average value and the awakening times of the user in each time period every day;
the adjustment module includes:
the first judgment module is used for judging whether the user daily behavior confidence coefficient average value is far higher than the preset confidence coefficient threshold value;
a first adjusting module, configured to integrally adjust the preset confidence level threshold downward when the user daily behavior confidence level average value is not much higher than the preset confidence level threshold;
the second judgment module is used for judging whether the time period of the current time is in a high-frequency use time period or a low-frequency use time period;
the second adjusting module is used for adjusting the preset confidence threshold value downwards when the current time period is in the high-frequency using time period, and adjusting the preset confidence threshold value upwards when the current time period is in the low-frequency using time period;
and the high-low frequency use time period determining module is used for determining the high-frequency use time period and the low-frequency use time period of each day of the user according to the awakening times of each time period of each day of the user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810516713.1A CN108847219B (en) | 2018-05-25 | 2018-05-25 | Awakening word preset confidence threshold adjusting method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810516713.1A CN108847219B (en) | 2018-05-25 | 2018-05-25 | Awakening word preset confidence threshold adjusting method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108847219A CN108847219A (en) | 2018-11-20 |
| CN108847219B true CN108847219B (en) | 2020-12-25 |
Family
ID=64213481
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810516713.1A Active CN108847219B (en) | 2018-05-25 | 2018-05-25 | Awakening word preset confidence threshold adjusting method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108847219B (en) |
Families Citing this family (30)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109256134B (en) * | 2018-11-22 | 2021-11-02 | 深圳市同行者科技有限公司 | Voice awakening method, storage medium and terminal |
| CN111261143B (en) * | 2018-12-03 | 2024-03-22 | 嘉楠明芯(北京)科技有限公司 | Voice wakeup method and device and computer readable storage medium |
| CN111292737A (en) * | 2018-12-07 | 2020-06-16 | 阿里巴巴集团控股有限公司 | Voice interaction and voice awakening detection method, device, equipment and storage medium |
| CN109859774B (en) * | 2019-01-02 | 2021-04-02 | 珠海格力电器股份有限公司 | Voice equipment and method and device for adjusting endpoint detection sensitivity thereof and storage medium |
| CN109920418B (en) * | 2019-02-20 | 2021-06-22 | 北京小米移动软件有限公司 | Method and device for adjusting awakening sensitivity |
| CN109801629A (en) * | 2019-03-01 | 2019-05-24 | 珠海格力电器股份有限公司 | A kind of sound control method, device, storage medium and air-conditioning |
| EP3709194A1 (en) | 2019-03-15 | 2020-09-16 | Spotify AB | Ensemble-based data comparison |
| CN110148405B (en) * | 2019-04-10 | 2021-07-13 | 北京梧桐车联科技有限责任公司 | Voice instruction processing method and device, electronic equipment and storage medium |
| CN109887507B (en) * | 2019-04-22 | 2021-03-12 | 成都启英泰伦科技有限公司 | Method for reducing similar voice command word misrecognition rate |
| CN110136707B (en) * | 2019-04-22 | 2021-03-02 | 云知声智能科技股份有限公司 | Man-machine interaction system for multi-equipment autonomous decision making |
| CN110211605A (en) * | 2019-05-24 | 2019-09-06 | 珠海多士科技有限公司 | Smart machine speech sensitivity adjusting method, device, equipment and storage medium |
| CN110047487B (en) * | 2019-06-05 | 2022-03-18 | 广州小鹏汽车科技有限公司 | Wake-up method and device for vehicle-mounted voice equipment, vehicle and machine-readable medium |
| KR20200141860A (en) | 2019-06-11 | 2020-12-21 | 삼성전자주식회사 | Electronic apparatus and the control method thereof |
| CN112086096B (en) * | 2019-06-14 | 2024-04-05 | 北京京东尚科信息技术有限公司 | Data processing method, device, system and medium |
| CN112102821B (en) * | 2019-06-18 | 2024-01-12 | 北京京东尚科信息技术有限公司 | Data processing method, device, system and medium applied to electronic equipment |
| CN110265018B (en) * | 2019-07-01 | 2022-03-04 | 成都启英泰伦科技有限公司 | Method for recognizing continuously-sent repeated command words |
| US11094319B2 (en) | 2019-08-30 | 2021-08-17 | Spotify Ab | Systems and methods for generating a cleaned version of ambient sound |
| US10827028B1 (en) | 2019-09-05 | 2020-11-03 | Spotify Ab | Systems and methods for playing media content on a target device |
| CN110570861B (en) * | 2019-09-24 | 2022-02-25 | Oppo广东移动通信有限公司 | Method and device for voice wake-up, terminal equipment and readable storage medium |
| CN110782891B (en) * | 2019-10-10 | 2022-02-18 | 珠海格力电器股份有限公司 | Audio processing method and device, computing equipment and storage medium |
| CN111128155B (en) * | 2019-12-05 | 2020-12-01 | 珠海格力电器股份有限公司 | Awakening method, device, equipment and medium for intelligent equipment |
| US11328722B2 (en) | 2020-02-11 | 2022-05-10 | Spotify Ab | Systems and methods for generating a singular voice audio stream |
| US11308959B2 (en) | 2020-02-11 | 2022-04-19 | Spotify Ab | Dynamic adjustment of wake word acceptance tolerance thresholds in voice-controlled devices |
| CN111427293A (en) * | 2020-03-26 | 2020-07-17 | 广州立功科技股份有限公司 | Multichannel input sampling awakening method and device and control equipment |
| CN111816178A (en) * | 2020-07-07 | 2020-10-23 | 云知声智能科技股份有限公司 | Voice equipment control method, device and equipment |
| CN114187904A (en) * | 2020-08-25 | 2022-03-15 | 广州华凌制冷设备有限公司 | Similarity threshold acquisition method, voice household appliance and computer readable storage medium |
| CN112767935B (en) * | 2020-12-28 | 2022-11-25 | 北京百度网讯科技有限公司 | Awakening index monitoring method and device and electronic equipment |
| CN112634897B (en) * | 2020-12-31 | 2022-10-28 | 青岛海尔科技有限公司 | Equipment awakening method and device, storage medium and electronic device |
| CN113488050B (en) * | 2021-07-09 | 2024-03-26 | 北京有竹居网络技术有限公司 | Voice wakeup method and device, storage medium and electronic equipment |
| CN113834192A (en) * | 2021-09-18 | 2021-12-24 | 珠海格力电器股份有限公司 | State adjustment method for voice recognition device in air conditioner and air conditioner |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011248107A (en) * | 2010-05-27 | 2011-12-08 | Nippon Telegr & Teleph Corp <Ntt> | Voice recognition result search method, apparatus and program for the same |
| CN103440867A (en) * | 2013-08-02 | 2013-12-11 | 安徽科大讯飞信息科技股份有限公司 | Method and system for recognizing voice |
| CN103578468A (en) * | 2012-08-01 | 2014-02-12 | 联想(北京)有限公司 | Method for adjusting confidence coefficient threshold of voice recognition and electronic device |
| CN105654949A (en) * | 2016-01-07 | 2016-06-08 | 北京云知声信息技术有限公司 | Voice wake-up method and device |
| CN106328148A (en) * | 2016-08-19 | 2017-01-11 | 上汽通用汽车有限公司 | Natural speech recognition method, natural speech recognition device and natural speech recognition system based on local and cloud hybrid recognition |
-
2018
- 2018-05-25 CN CN201810516713.1A patent/CN108847219B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011248107A (en) * | 2010-05-27 | 2011-12-08 | Nippon Telegr & Teleph Corp <Ntt> | Voice recognition result search method, apparatus and program for the same |
| CN103578468A (en) * | 2012-08-01 | 2014-02-12 | 联想(北京)有限公司 | Method for adjusting confidence coefficient threshold of voice recognition and electronic device |
| CN103440867A (en) * | 2013-08-02 | 2013-12-11 | 安徽科大讯飞信息科技股份有限公司 | Method and system for recognizing voice |
| CN105654949A (en) * | 2016-01-07 | 2016-06-08 | 北京云知声信息技术有限公司 | Voice wake-up method and device |
| CN106328148A (en) * | 2016-08-19 | 2017-01-11 | 上汽通用汽车有限公司 | Natural speech recognition method, natural speech recognition device and natural speech recognition system based on local and cloud hybrid recognition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108847219A (en) | 2018-11-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108847219B (en) | Awakening word preset confidence threshold adjusting method and system | |
| CN105654949B (en) | A kind of voice awakening method and device | |
| CN105118257B (en) | Intelligent control system and method | |
| CN108027952B (en) | Method and electronic device for providing content | |
| US9685161B2 (en) | Method for updating voiceprint feature model and terminal | |
| CN110176226A (en) | A kind of speech recognition and speech recognition modeling training method and device | |
| WO2015007092A1 (en) | Method, apparatus and device for controlling antenna of mobile device | |
| CN103886731B (en) | A kind of noise control method and equipment | |
| CN104064185B (en) | Information processing method and system, electronic equipment | |
| CN109473092B (en) | Voice endpoint detection method and device | |
| CN110070857B (en) | Model parameter adjusting method and device of voice awakening model and voice equipment | |
| CN108735209A (en) | Wake up word binding method, smart machine and storage medium | |
| CN109089156B (en) | Sound effect adjusting method and device and terminal | |
| CN111949323B (en) | Optimization method and device for waking up intelligent equipment, intelligent equipment and storage medium | |
| CN106203235B (en) | Living body identification method and apparatus | |
| CN112420041B (en) | Control method and device for wake-up threshold of intelligent device, intelligent device and storage medium | |
| CN204904540U (en) | There is screen intelligent remote controller | |
| CN111968644A (en) | Intelligent device awakening method and device and electronic device | |
| US20200134294A1 (en) | Method and Device for Face Recognition, Storage Medium, and Electronic Device | |
| CN104168263A (en) | Server and clock realization method thereof | |
| CN109124572A (en) | A kind of dormant judgment method, system and air purifier | |
| CN112185369A (en) | Volume adjusting method, device, equipment and medium based on voice control | |
| CN107508979B (en) | Volume adjusting method, device, terminal and storage medium | |
| CN204884159U (en) | There is not screen intelligent remote controller | |
| CN108810614A (en) | Method for regulation of sound volume, system and readable storage medium storing program for executing |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20190402 Address after: 201616 No. 3666 Sixian Road, Songjiang District, Shanghai Applicant after: PHICOMM (SHANGHAI) Co.,Ltd. Address before: 610000 No. 125 Longquan Street Park Road, Longquanyi District, Chengdu City, Sichuan Province Applicant before: SICHUAN FEIXUN QUANZHI INFORMATION TECHNOLOGY Co.,Ltd. |
|
| TA01 | Transfer of patent application right |
Effective date of registration: 20201109 Address after: Room 10242, No. 260, Jiangshu Road, Xixing street, Binjiang District, Hangzhou City, Zhejiang Province Applicant after: Hangzhou Jiji Intellectual Property Operation Co.,Ltd. Address before: 201616 Shanghai city Songjiang District Sixian Road No. 3666 Applicant before: Phicomm (Shanghai) Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20201208 Address after: Room 703, building 3, Yangguang Chuncheng, Lubei street, Luqiao District, Taizhou City, Zhejiang Province Applicant after: Taizhou zhiao communication equipment Co.,Ltd. Address before: Room 10242, No. 260, Jiangshu Road, Xixing street, Binjiang District, Hangzhou City, Zhejiang Province Applicant before: Hangzhou Jiji Intellectual Property Operation Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230109 Address after: 313000 room 1019, Xintiandi commercial office, Yishan street, Wuxing District, Huzhou, Zhejiang, China Patentee after: Huzhou YingLie Intellectual Property Operation Co.,Ltd. Address before: Room 703, building 3, Yangguang Chuncheng, Lubei street, Luqiao District, Taizhou City, Zhejiang Province Patentee before: Taizhou zhiao communication equipment Co.,Ltd. |