CA3164079A1 - Smart-device-orientated feedback awaking method and smart device thereof - Google Patents
Smart-device-orientated feedback awaking method and smart device thereof Download PDFInfo
- Publication number
- CA3164079A1 CA3164079A1 CA3164079A CA3164079A CA3164079A1 CA 3164079 A1 CA3164079 A1 CA 3164079A1 CA 3164079 A CA3164079 A CA 3164079A CA 3164079 A CA3164079 A CA 3164079A CA 3164079 A1 CA3164079 A1 CA 3164079A1
- Authority
- CA
- Canada
- Prior art keywords
- level
- feedback
- voice
- feedback mode
- reliability
- 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
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000009849 deactivation Effects 0.000 claims description 18
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000004044 response Effects 0.000 claims description 8
- 230000004913 activation Effects 0.000 claims description 6
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000002618 waking effect Effects 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/16—Sound input; Sound output
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/4401—Bootstrapping
- G06F9/4418—Suspend and resume; Hibernate and awake
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/02—Feature extraction for speech recognition; Selection of recognition unit
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; 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 OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L2015/223—Execution procedure of a spoken command
Abstract
Disclosed are a smart device wake-up and feedback method, and a smart device. The method comprises: detecting a voice command of a user, and extracting a voice feature of the voice command; matching the voice feature against a preset scoring index, and calculating the credibility of the voice command according to a matching result; comparing the credibility with a preset credibility threshold, and matching, according to a comparison result, the credibility to a feedback mode of a corresponding level; and implementing the feedback mode of the corresponding level. The hierarchical wake-up and feedback method and the smart device provided in the present invention avoid the situation in which direct voice feedback easily disturbs users, and improve the smart device use experience for users without preventing users from promptly obtaining feedback information of smart devices, thus providing voice feedback of smart devices in a flexible manner.
Description
SMART-DEVICE-ORIENTATED FEEDBACK AWAKING METHOD AND SMART
DEVICE THEREOF
BACKGROUND OF THE INVENTION
Technical Field [0001] The present invention relates to control of smart devices, and more particularly to a smart-device-orientated feedback awaking method and a smart device using the method.
Description of Related Art
DEVICE THEREOF
BACKGROUND OF THE INVENTION
Technical Field [0001] The present invention relates to control of smart devices, and more particularly to a smart-device-orientated feedback awaking method and a smart device using the method.
Description of Related Art
[0002] Voice control is currently a popular and dominant way to control smart devices due to its advantages about easy operation and flexible use. A user can simply input voice instructions to a smart device and receive feedbacks from the smart device.
When waked by voice, the smart device immediately outputs a voice message to prompt the user to input further instructions. However, if the user wakes the smart device unintentionally, the smart device may perform operations wrongly. Besides, sudden voice feedbacks in response to unintentional waking voice can disturb the user, leading to poor user experience.
SUMMARY OF THE INVENTION
When waked by voice, the smart device immediately outputs a voice message to prompt the user to input further instructions. However, if the user wakes the smart device unintentionally, the smart device may perform operations wrongly. Besides, sudden voice feedbacks in response to unintentional waking voice can disturb the user, leading to poor user experience.
SUMMARY OF THE INVENTION
[0003] In order to address the shortcomings of the prior art, embodiments of the present invention provide a smart-device-orientated rank-based feedback awaking method a smart device.
T technical schemes are as below.
T technical schemes are as below.
[0004] In one aspect, the present invention provides a smart-device-orientated feedback awaking method, which comprises:
[0005] hearing a voice instruction from a user, and extracting voice features of the voice Date Recue/Date Received 2022-06-08 instruction;
[0006] matching the voice features with preset rating indicators, and determining a reliability level of the voice instruction through computation according to matching results;
[0007] comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result; and
[0008] executing the feedback mode of the corresponding level.
[0009] Further, the step of comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result comprises:
[0010] if reliability level is not smaller than a first threshold, assigning a Level-Three feedback mode;
[0011] if the reliability level is smaller than the first threshold and not smaller than a second threshold, assigning a Level-Two feedback mode; and
[0012] if the reliability level is smaller than the second threshold, assigning a Level-One feedback mode.
[0013] Further, the Level-One feedback mode comprises a silence mode; the Level-Two feedback mode comprises: an indicator lamp feedback or a vibration feedback;
and the Level-Three feedback mode comprises: a voice feedback.
and the Level-Three feedback mode comprises: a voice feedback.
[0014] Further, the method further comprises: counting a waiting duration since activation of the feedback mode, if the waiting duration is greater than predetermined waiting duration, automatically going silence and deactivation, and if the waiting duration is not greater than the predetermined waiting duration, going silence in response to a deactivation instruction from the user.
[0015] Further, the step of extracting voice features of the voice instruction at least comprises:
determining the semantic similarity between the voice instruction and a predetermined standard awaking instruction;
determining the semantic similarity between the voice instruction and a predetermined standard awaking instruction;
[0016] and comprises any one or two of:
[0017] determining a sound intensity level of the voice instruction; and Date Recue/Date Received 2022-06-08
[0018] recognizing plural keywords in the voice instruction according to plural keywords in the predetermined standard awaking instruction, determining a time interval between the keywords in the voice instruction, and matching the time interval between the keywords in the voice instruction to a preset time interval criterion, so as to obtain a continuity index;
[0019] In which the step of determining a reliability level of the voice instruction through computation comprises:
[0020] performing a weighted average calculation on the matching results of the voice features so as to obtain the reliability level.
[0021] In another aspect, the present invention provides a smart device, which comprises:
[0022] a hearing module, for hearing a voice instruction from a user;
[0023] a feature extracting module, for extracting voice features of the voice instruction;
[0024] a reliability computation module, for matching the voice features with preset rating indicators, and determining a reliability level of the voice instruction through computation according to matching results;
[0025] a feedback mode determining module, for comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result; and
[0026] an execution module, for executing the feedback mode of the corresponding level.
[0027] Further, the feedback mode determining module comprises:
[0028] a Level-Three feedback mode assigning module, for assigning a Level-Three feedback mode for the voice instruction whose reliability level is not smaller than first threshold;
[0029] a Level-Two feedback mode assigning module, for assigning a Level-Two feedback mode for the voice instruction whose reliability level is smaller than the first threshold and not smaller than a second threshold; and
[0030] a Level-One feedback mode assigning module, for assigning a Level-One feedback mode for the voice instruction whose reliability level is smaller than the second threshold.
[0031] Further, the Level-One feedback mode comprises a silence mode; the Level-Two feedback mode comprises: an indicator lamp feedback or a vibration feedback;
the Level-Date Recue/Date Received 2022-06-08 Three feedback mode comprises: voice feedback.
the Level-Date Recue/Date Received 2022-06-08 Three feedback mode comprises: voice feedback.
[0032] The execution module comprises: an indicator lamp and/or a vibration apparatus and a voice apparatus.
[0033] Further, the smart device further comprises a deactivation module, for counting a waiting duration since activation of the feedback mode, if the waiting duration is greater than predetermined waiting duration, automatically going silence and deactivation, and if the waiting duration is not greater than the redetermined waiting duration, going silence in response to a deactivation instruction from the user.
[0034] Further, the feature extracting module comprises: a semantic similarity computation module, for determining the semantic similarity between the voice instruction and a predetermined standard awaking instruction;
[0035] and comprises one or two of feature computation modules, including:
[0036] a sound intensity recognizing module, for determining a sound intensity level of the voice instruction; and
[0037] a continuity index computation module, for determining plural keywords in the voice instruction according to plural keywords in the predetermined standard awaking instruction, determining a time interval between the keywords in the voice instruction, and matching the time interval between the keywords in the voice instruction to a preset time interval criterion, so as to obtain a continuity index;
[0038] Preferably, when the feature extracting module comprises plural feature computation modules, the reliability computation module performs a weighted average calculation on the matching results of the voice features so as to obtain the reliability level.
[0039] The technical schemes of embodiments of the present invention provide the following beneficial effects:
[0040] The rank-based feedback awaking method and the smart device improve user experience by eliminating unnecessary disturbance to users without preventing users from receiving feedbacks from the smart device timely, thereby making voice feedbacks of the smart device more flexible.
Date Recue/Date Received 2022-06-08 BRIEF DESCRIPTION OF THE DRAWINGS
Date Recue/Date Received 2022-06-08 BRIEF DESCRIPTION OF THE DRAWINGS
[0041] To better illustrate the technical schemes as disclosed in the embodiments of the present invention, accompanying drawings referred in the description of the embodiments below are introduced briefly. It is apparent that the accompanying drawings as recited in the following description merely provide a part of possible embodiments of the present invention, and people of ordinary skill in the art would be able to obtain more drawings according to those provided herein without paying creative efforts, wherein:
[0042] FIG. 1 is a flowchart of a smart-device-orientated rank-based feedback awaking method according to a first embodiment of the present invention;
[0043] FIG. 2 is a modular block diagram of a smart device according to a second embodiment of the present invention; and
[0044] FIG. 3 is a flowchart of a smart-speaker-orientated rank-based feedback awaking method according to a third embodiment of the present invention.
DETAILED DESCRIPTION OF THE INVENTION
DETAILED DESCRIPTION OF THE INVENTION
[0045] To make the foregoing objectives, features, and advantages of the present invention clearer and more understandable, the following description will be directed to some embodiments as depicted in the accompanying drawings to detail the technical schemes disclosed in these embodiments. It is, however, to be understood that the embodiments referred herein are only a part of all possible embodiments and thus not exhaustive. Based on the embodiments of the present invention, all the other embodiments can be conceived without creative labor by people of ordinary skill in the art, and all these and other embodiments shall be encompassed in the scope of the present invention.
[0046] In the prior art, when a smart device is waked up, it usually gives a voice feedback immediately. If the awaking voice is made by a user unintentionally, the sudden voice feedback provided by the smart device can disturb the user, leading to poor user experience. Therefore, in order to prevent an unexpected voice feedback made by a smart Date Recue/Date Received 2022-06-08 device from disturbing a user, one embodiments of the present invention provide a rank-based method for giving feedbacks. The method assigns feedback mode according to voice features in voice instructions from users, so as to gives feedbacks with different influence levels. In this way, it is possible to for the smart device tell whether a user awakes a smart device on purpose or not, thereby preventing unexpected voice feedbacks from disturbing users while allowing users to operate devices normally. The specific technical schemes are described with reference to some embodiments.
[0047] Embodiment 1
[0048] As shown in FIG. 1, a smart-device-orientated feedback awaking method, comprises:
[0049] hearing a voice instruction from a user, and extracting voice features of the voice instruction;
[0050] matching the voice features with preset rating indicators, and determining a reliability level of the voice instruction through computation according to matching results;
[0051] comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result; and
[0052] executing the feedback mode of the corresponding level.
[0053] A smart device executing the foregoing method stores correspondence between thresholds and scores for different voice features. Thus, the acquired voice features can be directly mapped to their corresponding scores. According to the scores, the reliability level of the voice instruction can be determined. When the voice features comprise multiple dimensions, weighted calculation is performed on the scores of the voice features in different dimensions, so as to determine the reliability level of the voice instruction.
Afterward, the reliability level is compared with a predetermined threshold preloaded in the smart device. The smart device stores correspondence between reliability threshold ranges and feedback modes. When it is ascertained that in which threshold range the score of the voice instruction is in, the current smart device executes the feedback mode corresponding to the threshold range, thereby matching the feedback modes to different voice instructions.
Date Recue/Date Received 2022-06-08
Afterward, the reliability level is compared with a predetermined threshold preloaded in the smart device. The smart device stores correspondence between reliability threshold ranges and feedback modes. When it is ascertained that in which threshold range the score of the voice instruction is in, the current smart device executes the feedback mode corresponding to the threshold range, thereby matching the feedback modes to different voice instructions.
Date Recue/Date Received 2022-06-08
[0054] In the foregoing method, for computing the voice features of the voice instruction, one or more of dimensions of features, including semantic similarity, sound intensity levels, the continuity index can be selected. The step of extracting voice features of the voice instruction comprises: determining the semantic similarity between the voice instruction and a predetermined standard waking instruction;
[0055] and comprises any one or two of:
[0056] determining a sound intensity level of the voice instruction; and
[0057] recognizing plural keywords in the voice instruction according to plural keywords in the predetermined standard awaking instruction, determining a time interval between the keywords in the voice instruction, and matching the time interval between the keywords in the voice instruction to a preset time interval criterion, so as to obtain a continuity index.
[0058] If there are multiple types of voice features, the step of determining the reliability levels of the voice instruction comprises:
[0059] performing a weighted average calculation on the matching results of the voice features so as to obtain the reliability level. To this end, the exact weights are determined according to significance levels of the voice features.
[0060] In the disclosed method, semantic similarity between of the voice instruction and a predetermined standard awaking instruction is determined to make sure whether the voice instruction is an awaking instruction, thereby identifying intention of a user awaking the smart device. Computation of semantic similarity may be achieved through:
dividing the voice instruction into several keywords, comparing the keywords from the voice instruction with keywords in the predetermined standard awaking instruction, counting a number of common keywords, determining a ratio of the number to a total number of the keywords in the predetermined standard awaking instruction, and determining the semantic similarity according to the ratio. Alternatively, for a complicated voice instruction, a neural network model may be used to compare the keywords in the voice instruction with the keywords of the predetermined standard voice instruction to compute semantic similarity smartly.
Date Recue/Date Received 2022-06-08
dividing the voice instruction into several keywords, comparing the keywords from the voice instruction with keywords in the predetermined standard awaking instruction, counting a number of common keywords, determining a ratio of the number to a total number of the keywords in the predetermined standard awaking instruction, and determining the semantic similarity according to the ratio. Alternatively, for a complicated voice instruction, a neural network model may be used to compare the keywords in the voice instruction with the keywords of the predetermined standard voice instruction to compute semantic similarity smartly.
Date Recue/Date Received 2022-06-08
[0061] The sound intensity of the voice instruction is mainly used to further determine the intention of the user awaking the smart device. If the sound intensity is high, it is determined that the user was sending the voice instruction to the smart device intentionally. Specifically, this can be achieved using a volume test module in the smart device.
[0062] The continuity index of the voice instruction may be used to further determine whether the voice instruction from the user is an awaking instruction. If there are many common keywords between the voice instruction and the standard awaking instruction, but the time interval between two keywords in the voice instruction is long, it can be determined that the voice instruction is with less intention of awaking the smart device.
[0063] In the method, for differentiating feedback ranks and decreasing computation loads of the smart device, the feedback modes preferably including a Level-One feedback mode, a Level-Two feedback mode, and a Level-Three feedback mode. The three feedback modes of different ranks may give feedbacks in the same way or in different ways.
The three feedback modes of different ranks represent different influence levels on hearing and/or sight of users. As shown in FIG. 1, the step of comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result comprises:
The three feedback modes of different ranks represent different influence levels on hearing and/or sight of users. As shown in FIG. 1, the step of comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result comprises:
[0064] if reliability level is not smaller than a first threshold, assigning a Level-Three feedback mode;
[0065] if the reliability level is smaller than the first threshold and not smaller than a second threshold, assigning a Level-Two feedback mode; and
[0066] if the reliability level is smaller than the second threshold, assigning a Level-One feedback mode.
[0067] In the disclosed method, the smart device decides which feedback mode is to be activated according to the voice instruction from the user. The feedback modes of three different levels are configured to give feedbacks differently. In particular, the Level-One feedback mode comprises a silence mode. The Level-Two feedback mode gives indicator lamp feedbacks or vibration feedbacks. The Level-Three feedback mode gives voice feedbacks.
Date Recue/Date Received 2022-06-08 If the smart device assigns the silence mode for the current voice instruction, the smart device wakes but does not give any feedback. If smart device decides that it has to give an indicator lamp feedback or a vibration feedback, the smart device flashes an indicator lamp or generates vibration to answer to the awaking instruction. In practical use, whether the feedback is given as light of an indicator lamp or vibration is determined by the manufacturer according to the hardware conditions of the smart device.
Date Recue/Date Received 2022-06-08 If the smart device assigns the silence mode for the current voice instruction, the smart device wakes but does not give any feedback. If smart device decides that it has to give an indicator lamp feedback or a vibration feedback, the smart device flashes an indicator lamp or generates vibration to answer to the awaking instruction. In practical use, whether the feedback is given as light of an indicator lamp or vibration is determined by the manufacturer according to the hardware conditions of the smart device.
[0068] After the smart device executes the determined feedback mode, the waiting duration before the next instruction is received is counted. If the waiting duration is greater than a predetermined waiting duration, the smart device automatically goes silence and deactivation. If the waiting duration is not greater than the predetermined waiting duration, the smart device goes silence in response to a deactivation instruction from the user.
[0069] As shown in FIG. 2, based on the smart-device-orientated rank-based feedback awaking method as described previously, embodiments of the present invention further provide a smart device having a rank-based awaking function. The smart device comprises:
[0070] a hearing module, for hearing a voice instruction from a user;
[0071] feature extracting module extracting voice features of the voice instruction;
[0072] a reliability computation module, for matching the voice features with preset rating indicators, and determining a reliability level of the voice instruction through computation according to matching results;
[0073] a feedback mode determining module, for comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result;
[0074] an execution module, for executing feedback modes corresponding different ranks; and
[0075] a deactivation module, for counting a waiting duration since activation of the feedback mode, if the waiting duration is greater than predetermined waiting duration, automatically going silence and deactivation, and if the waiting duration is not greater than the redetermined waiting duration, going silence in response to a deactivation instruction from the user.
[0076] Therein, the feature extracting module comprises one or more of the following feature Date Recue/Date Received 2022-06-08 computation modules:
[0077] a semantic similarity computation module, for determining the semantic similarity between the voice instruction and a predetermined standard awaking instruction;
[0078] a sound intensity recognizing module, for determining a sound intensity level of the voice instruction; and
[0079] a continuity index computation module, for determining plural keywords in the voice instruction according to plural keywords in the predetermined standard awaking instruction, determining a time interval between the keywords in the voice instruction, and matching the time interval between the keywords in the voice instruction to a preset time interval criterion, so as to obtain a continuity index;
[0080] When the feature extracting module comprises multiple feature computation modules, the reliability computation module performs weighted average calculation on the scores of the voice features so as to obtain the reliability level.
[0081] Therein, for achieving rank-based feedback, the feedback mode determining module comprises:
[0082] a Level-Three feedback mode confirming module, for assigning a Level-Three feedback mode for the voice instruction whose reliability level is not smaller than first threshold;
[0083] Level-Two feedback mode confirming module, for assigning a Level-Two feedback mode for the voice instruction whose reliability level is smaller than the first threshold and not smaller than a second threshold; and
[0084] Level-One feedback mode confirming module, for assigning a Level-One feedback mode for the voice instruction whose reliability level is smaller than the second threshold.
[0085] Therein, the Level-Three feedback mode generates feedbacks having an influence on sight and/or hearing users greater than that of the feedbacks of the Level-Two feedback mode, and the Level-Two feedback mode generates feedbacks having the influence on sight and/or hearing users greater than that of the feedbacks of the first feedback mode.
The feedback modes of the three levels may give their feedbacks in the same way or in different ways.
The feedback modes of the three levels may give their feedbacks in the same way or in different ways.
[0086] In the three feedback modes, the Level-One feedback mode is a silence mode. The Level-Date Recue/Date Received 2022-06-08 Two feedback mode gives indicator lamp feedbacks or vibration feedbacks. The Level-Three feedback mode comprises: a voice feedback. Correspondingly, the execution module comprises: an indicator lamp, a vibration apparatus, and a voice apparatus.
[0087] In operation of the smart device, the hearing module hears a voice instruction, and sends the voice instruction to the feature extracting module for extraction of features. After the features are extracted, the extracted features are set to the reliability computation module for determining the reliability of the voice instruction. The feedback mode determines which feedback mode the smart device will use according to the reliability level, and signals execution apparatuses in the execution module to provide rank-based feedbacks.
[0088] Embodiment 2
[0089] The present embodiment is based on Embodiment 1 and is for further explaining the present invention with reference to a practical application scenario.
[0090] As shown in FIG. 3, an awaking method for a smart speaker comprises the following steps.
[0091] At SOL a voice instruction from a user is heard.
[0092] At S02, voice features of the voice instruction are extracted. The voice features include:
semantic similarity, sound intensity, and a continuity index.
semantic similarity, sound intensity, and a continuity index.
[0093] S03, the foregoing voice features are matched to preset rating indicators so as to obtain scores of the voice features.
[0094] For example, the matching result of semantic similarity is 3, the matching result of sound intensity is 2, and the matching result of the continuity index is 4.
[0095] At SO4, weighted calculation is performed on the scores of the voice features, so as to obtain the reliability level of the voice instruction.
[0096] For example, the reliability level may be calculated using the equation below:
[0097] Reliability =0.4xMatching Result of Semantic Similarity + 0.3 xMatching Result of Sound Intensity + 0.3 xMatching Result of Continuity index
[0098] Based on the discussed example, the reliability level is 3.
[0099] At 505, the reliability level is compared with a preset reliability threshold, and a feedback mode of a corresponding level matching the reliability level is assigned according to the Date Recue/Date Received 2022-06-08 comparison result.
[0100] For example, as set in the smart speaker, the first threshold is 5, the second threshold is 3.
[0101] If the reliability level is not smaller than a first threshold, a Level-Three feedback mode is assigned.
[0102] If the reliability level is smaller than the first threshold and not smaller than a second threshold, a Level-Two feedback mode is assigned.
[0103] If the reliability level is smaller than the second threshold, a Level-One feedback mode is assigned.
[0104] The reliability level of the voice instruction is compared with the foregoing threshold, and for this reliability level, the Level-Two feedback mode is assigned.
[0105] At S06, the feedback mode of the corresponding level is executed.
[0106] Based on the discussed example, in smart speaker, the Level-One feedback mode is a silence mode. The Level-Two feedback mode is an indicator lamp feedback mode.
The Level-Three feedback mode is a voice feedback mode. If the reliability level of the voice instruction matches the Level-Two feedback mode, the smart speaker flashes its indicator lamp to prompt the user that the smart speaker now wakes and the user can give his/her next instruction.
The Level-Three feedback mode is a voice feedback mode. If the reliability level of the voice instruction matches the Level-Two feedback mode, the smart speaker flashes its indicator lamp to prompt the user that the smart speaker now wakes and the user can give his/her next instruction.
[0107] Based on the foregoing smart-speaker-orientated rank-based feedback awaking method, the smart speaker of the present embodiment comprises the following components.
[0108] The hearing module is for hearing a voice instruction from a user.
[0109] The feature extracting module is for extracting voice features in the voice instruction.
[0110] The feature extracting module comprises: a semantic similarity computation module, for determining the semantic similarity between the voice instruction and a predetermined standard awaking instruction; a sound intensity recognizing module, for determining a sound intensity level of the voice instruction; and a continuity index computation module, for determining a time interval between the keywords in the voice instruction, matching a time interval between the keywords in the voice instruction to a preset time interval criterion, so as to obtain a continuity index; a reliability computation module, for Date Recue/Date Received 2022-06-08 performing weighted average calculation on the scores of the voice features so as to obtain the reliability level.
[0111] a reliability computation module, for matching the voice features with preset rating indicators, and determining a reliability level of the voice instruction through computation according to matching results.
[0112] a feedback mode determining module, for comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result.
[0113] an execution module, for executing the feedback mode of the corresponding level.
feedback mode determining module comprises the following modules.
feedback mode determining module comprises the following modules.
[0114] The Level-Three feedback mode confirming module is for assigning a Level-Three feedback mode for a voice instruction whose reliability level is not smaller than a first threshold. Level-Two feedback mode confirming module is for assigning a Level-Two feedback mode for the voice instruction whose reliability level is smaller than the first threshold and not smaller than a second threshold; and Level-One feedback mode confirming module is for assigning a Level-One feedback mode for the voice instruction whose reliability level is smaller than the second threshold.
[0115] The deactivation module is for counting a waiting duration since activation of the feedback mode, if the waiting duration is greater than predetermined waiting duration, automatically going silence and deactivation, and if the waiting duration is not greater than the redetermined waiting duration, going silence in response to a deactivation instruction from the user.
[0116] The voice apparatus is for executing the Level-Three feedback mode, and gives a voice feedback accordingly.
[0117] The indicator lamp is for executing the Level-Two feedback mode, and gives prompt using the indicator lamp.
[0118] The technical schemes of embodiments of the present invention provide the following beneficial effects:
[0119] The rank-based feedback awaking method and the smart device of the present invention Date Recue/Date Received 2022-06-08 improve user experience by eliminating unnecessary disturbance to users without preventing users from receiving feedbacks from the smart device timely, thereby making voice feedbacks of the smart device more flexible.
[0120] All of the foregoing alternative technical scheme may be combined in an arbitrary manner to form an alternative embodiment of the present invention, and these possible combinations are not exhaustively described herein.
[0121] The present invention has been described with reference to the preferred embodiments and it is understood that the embodiments are not intended to limit the scope of the present invention. Moreover, as the contents disclosed herein should be readily understood and can be implemented by a person skilled in the art, all equivalent changes or modifications which do not depart from the concept of the present invention should be encompassed by the appended claims. Hence, the scope of the present invention shall only be defined by the appended claims.
Date Recue/Date Received 2022-06-08
Date Recue/Date Received 2022-06-08
Claims (10)
1. A smart-device-orientated feedback awaking method, comprising:
hearing a voice instruction from a user, and extracting voice features of the voice instruction;
matching the voice features with preset rating indicators, and determining a reliability level of the voice instruction through computation according to matching results;
wherein the voice features include: a sound intensity level, a semantic similarity, and a continuity index; wherein the continuity index is obtained by matching the time interval between the keywords in the voice instruction to a preset time interval criterion, while the reliability level is obtained by performing a weighted average calculation on the rating of the voice features;
comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result; if the reliability threshold is larger, the feedback mode of a corresponding level is larger, and if feedback mode of a corresponding level is larger, the influence on the user's hearing and/or vision is larger; and executing the feedback mode of the corresponding level.
hearing a voice instruction from a user, and extracting voice features of the voice instruction;
matching the voice features with preset rating indicators, and determining a reliability level of the voice instruction through computation according to matching results;
wherein the voice features include: a sound intensity level, a semantic similarity, and a continuity index; wherein the continuity index is obtained by matching the time interval between the keywords in the voice instruction to a preset time interval criterion, while the reliability level is obtained by performing a weighted average calculation on the rating of the voice features;
comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result; if the reliability threshold is larger, the feedback mode of a corresponding level is larger, and if feedback mode of a corresponding level is larger, the influence on the user's hearing and/or vision is larger; and executing the feedback mode of the corresponding level.
2. The smart-device-orientated feedback awaking method of claim 1, wherein the step of comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result comprises:
if reliability level is not smaller than a first threshold, assigning a Level-Three feedback mode;
if the reliability level is smaller than the first threshold and not smaller than a second threshold, assigning a Level-Two feedback mode; and if the reliability level is smaller than the second threshold, assigning a Level-One feedback mode.
if reliability level is not smaller than a first threshold, assigning a Level-Three feedback mode;
if the reliability level is smaller than the first threshold and not smaller than a second threshold, assigning a Level-Two feedback mode; and if the reliability level is smaller than the second threshold, assigning a Level-One feedback mode.
3. The smart-device-orientated feedback awaking method of claim 2, wherein the Level-One feedback mode comprises a silence mode; the Level-Two feedback mode comprises:
an indicator lamp feedback or a vibration feedback; and the Level-Three feedback mode comprises: a voice feedback.
an indicator lamp feedback or a vibration feedback; and the Level-Three feedback mode comprises: a voice feedback.
4. The smart-device-orientated feedback awaking method of claim 1, further comprising:
counting a waiting duration since activation of the feedback mode, if the waiting duration is greater than predetermined waiting duration, automatically going silence and deactivation, and if the waiting duration is not greater than the predetermined waiting duration, going silence in response to a deactivation instruction from the user.
counting a waiting duration since activation of the feedback mode, if the waiting duration is greater than predetermined waiting duration, automatically going silence and deactivation, and if the waiting duration is not greater than the predetermined waiting duration, going silence in response to a deactivation instruction from the user.
5. The smart-device-orientated feedback awaking method of any of claims 1 through 4, wherein the step of extracting voice features of the voice instruction comprises:
determining the semantic similarity between the voice instruction and a predetermined standard awaking instruction;
determining a sound intensity level of the voice instruction; and determining plural keywords in the voice instruction according to plural keywords in the predetermined standard awaking instruction, determining a time interval between the keywords in the voice instruction, and matching the time interval between the keywords in the voice instruction to a preset time interval criterion, so as to obtain a continuity index.
determining the semantic similarity between the voice instruction and a predetermined standard awaking instruction;
determining a sound intensity level of the voice instruction; and determining plural keywords in the voice instruction according to plural keywords in the predetermined standard awaking instruction, determining a time interval between the keywords in the voice instruction, and matching the time interval between the keywords in the voice instruction to a preset time interval criterion, so as to obtain a continuity index.
6. A smart device, comprising:
a hearing module, for hearing a voice instruction from a user;
a feature extracting module, for extracting voice features of the voice instruction;
a reliability computation module, for matching the voice features with preset rating indicators, and determining a reliability level of the voice instruction through computation according to matching results; wherein the voice features include: a sound intensity level, a semantic similarity, and a continuity index; wherein the continuity index is obtained by matching the time interval between the keywords in the voice instruction to a preset time interval criterion, while the reliability level is obtained by performing a weighted average calculation on the rating of the voice features;
a feedback mode determining module, for comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result; if the reliability threshold is larger, the feedback mode of a corresponding level is larger, and if feedback mode of a corresponding level is larger, the influence on the user's hearing and/or vision is larger; and an execution module, for executing the feedback mode of the corresponding level.
a hearing module, for hearing a voice instruction from a user;
a feature extracting module, for extracting voice features of the voice instruction;
a reliability computation module, for matching the voice features with preset rating indicators, and determining a reliability level of the voice instruction through computation according to matching results; wherein the voice features include: a sound intensity level, a semantic similarity, and a continuity index; wherein the continuity index is obtained by matching the time interval between the keywords in the voice instruction to a preset time interval criterion, while the reliability level is obtained by performing a weighted average calculation on the rating of the voice features;
a feedback mode determining module, for comparing the reliability level to a preset reliability threshold, and assigning a feedback mode of a corresponding level matching the reliability level according to a comparison result; if the reliability threshold is larger, the feedback mode of a corresponding level is larger, and if feedback mode of a corresponding level is larger, the influence on the user's hearing and/or vision is larger; and an execution module, for executing the feedback mode of the corresponding level.
7. The smart device of claim 6, wherein the feedback mode determining module comprises:
a Level-Three feedback mode assigning module, for assigning a Level-Three feedback mode for the voice instruction whose reliability level is not smaller than a first threshold;
a Level-Two feedback mode assigning module, for assigning a Level-Two feedback mode for the voice instruction whose reliability level is smaller than the first threshold and is not smaller than a second mode; and a Level-One feedback mode assigning module, for assigning a Level-One feedback mode of the voice instruction whose reliability level is smaller than the second threshold.
a Level-Three feedback mode assigning module, for assigning a Level-Three feedback mode for the voice instruction whose reliability level is not smaller than a first threshold;
a Level-Two feedback mode assigning module, for assigning a Level-Two feedback mode for the voice instruction whose reliability level is smaller than the first threshold and is not smaller than a second mode; and a Level-One feedback mode assigning module, for assigning a Level-One feedback mode of the voice instruction whose reliability level is smaller than the second threshold.
8. The smart device of claim 7, wherein the Level-One feedback mode comprises a silence mode;
the Level-Two feedback mode comprises: indicator lamp feedback and/or vibration feedback;
the Level-Three feedback mode comprises: voice feedback;
the execution module comprises: indicator lamp and/or vibration apparatus, and voice apparatus.
the Level-Two feedback mode comprises: indicator lamp feedback and/or vibration feedback;
the Level-Three feedback mode comprises: voice feedback;
the execution module comprises: indicator lamp and/or vibration apparatus, and voice apparatus.
9. The smart device of claim 6, further comprising a deactivation module, for counting a waiting duration since activation of the feedback mode, if the waiting duration is greater than predetermined waiting duration, automatically going silence and deactivation, and if the waiting duration is not greater than the redetermined waiting duration, going silence in response to a deactivation instruction from the user.
10. The smart device of any of claims 6 through 9, wherein the feature extracting module comprises:
a semantic similarity computation module, for determining the semantic similarity between the voice instruction and a predetermined standard awaking instruction;
a sound intensity recognizing module, for determining a sound intensity level of the voice instruction; and a continuity index computation module, for determining plural keywords in the voice instruction according to plural keywords in the predetermined standard awaking instruction, determining a time interval between the keywords in the voice instruction, and matching the time interval between the keywords in the voice instruction to a preset time interval criterion, so as to obtain a continuity index.
a semantic similarity computation module, for determining the semantic similarity between the voice instruction and a predetermined standard awaking instruction;
a sound intensity recognizing module, for determining a sound intensity level of the voice instruction; and a continuity index computation module, for determining plural keywords in the voice instruction according to plural keywords in the predetermined standard awaking instruction, determining a time interval between the keywords in the voice instruction, and matching the time interval between the keywords in the voice instruction to a preset time interval criterion, so as to obtain a continuity index.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911247176.6 | 2019-12-09 | ||
| CN201911247176.6A CN110673821B (en) | 2019-12-09 | 2019-12-09 | Intelligent device awakening feedback method and intelligent device |
| PCT/CN2020/105956 WO2021114687A1 (en) | 2019-12-09 | 2020-07-30 | Smart device wake-up and feedback method, and smart device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA3164079A1 true CA3164079A1 (en) | 2021-06-17 |
Family
ID=69088364
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA3164079A Pending CA3164079A1 (en) | 2019-12-09 | 2020-07-30 | Smart-device-orientated feedback awaking method and smart device thereof |
Country Status (3)
| Country | Link |
|---|---|
| CN (1) | CN110673821B (en) |
| CA (1) | CA3164079A1 (en) |
| WO (1) | WO2021114687A1 (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110673821B (en) * | 2019-12-09 | 2020-05-01 | 苏宁云计算有限公司 | Intelligent device awakening feedback method and intelligent device |
| CN111429901B (en) * | 2020-03-16 | 2023-03-21 | 云知声智能科技股份有限公司 | IoT chip-oriented multi-stage voice intelligent awakening method and system |
| CN114875622B (en) * | 2022-05-31 | 2023-08-15 | 四川虹美智能科技有限公司 | Intelligent airing control method and device |
| CN115691497B (en) * | 2023-01-04 | 2023-03-31 | 深圳市大晶光电科技有限公司 | Voice control method, device, equipment and medium |
| CN116562271B (en) * | 2023-07-10 | 2023-10-10 | 之江实验室 | Quality control method and device for electronic medical record, storage medium and electronic equipment |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20150025646A (en) * | 2013-08-29 | 2015-03-11 | 삼성전자주식회사 | Method for generating sound source and electronic device thereof |
| CN109391528A (en) * | 2018-08-31 | 2019-02-26 | 百度在线网络技术（北京）有限公司 | Awakening method, device, equipment and the storage medium of speech-sound intelligent equipment |
| CN109166580A (en) * | 2018-09-17 | 2019-01-08 | 珠海格力电器股份有限公司 | A kind of voice feedback prompt control method, system and air conditioner |
| 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 |
| CN110349579B (en) * | 2019-07-15 | 2021-10-15 | 北京梧桐车联科技有限责任公司 | Voice wake-up processing method and device, electronic equipment and storage medium |
| CN110503952B (en) * | 2019-07-29 | 2022-02-22 | 北京搜狗科技发展有限公司 | Voice processing method and device and electronic equipment |
| CN110673821B (en) * | 2019-12-09 | 2020-05-01 | 苏宁云计算有限公司 | Intelligent device awakening feedback method and intelligent device |
-
2019
- 2019-12-09 CN CN201911247176.6A patent/CN110673821B/en active Active
-
2020
- 2020-07-30 CA CA3164079A patent/CA3164079A1/en active Pending
- 2020-07-30 WO PCT/CN2020/105956 patent/WO2021114687A1/en active Application Filing
Also Published As
| Publication number | Publication date |
|---|---|
| WO2021114687A1 (en) | 2021-06-17 |
| CN110673821B (en) | 2020-05-01 |
| CN110673821A (en) | 2020-01-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA3164079A1 (en) | Smart-device-orientated feedback awaking method and smart device thereof | |
| US11355107B2 (en) | Voice smart device wake-up method, apparatus, device and storage medium | |
| CN109671426B (en) | Voice control method and device, storage medium and air conditioner | |
| US20230041517A1 (en) | Selectively activating on-device speech recognition, and using recognized text in selectively activating on-device nlu and/or on-device fulfillment | |
| US20180285068A1 (en) | Processing method of audio control and electronic device thereof | |
| US11741953B2 (en) | Using corrections, of automated assistant functions, for training of on-device machine learning models | |
| US20200211545A1 (en) | Voice interaction method, apparatus and device, and storage medium | |
| CN109686368B (en) | Voice wake-up response processing method and device, electronic equipment and storage medium | |
| CN110349579B (en) | Voice wake-up processing method and device, electronic equipment and storage medium | |
| KR20190109916A (en) | A electronic apparatus and a server for processing received data from the apparatus | |
| EP3956884A1 (en) | Identification and utilization of misrecognitions in automatic speech recognition | |
| US11361764B1 (en) | Device naming-indicator generation | |
| KR20230020523A (en) | Automatic hotword threshold tuning | |
| JP2022519648A (en) | Voice query service quality of service based on content metadata calculated by the client | |
| CN115552515A (en) | User intermediary for hotword/keyword detection | |
| KR20230104712A (en) | Hotword recognition adaptation based on personalized negatives | |
| KR20200007530A (en) | Method for processing user voice input and electronic device supporting the same | |
| CN113255362B (en) | Method and device for filtering and identifying human voice, electronic device and storage medium | |
| US11830486B2 (en) | Detecting near matches to a hotword or phrase | |
| KR20190114325A (en) | The apparatus for processing user voice input | |
| CN106156022B (en) | Information processing method and electronic equipment | |
| US20230143177A1 (en) | Contextual suppression of assistant command(s) | |
| CN108833688B (en) | Position reminding method and device, storage medium and electronic equipment | |
| CN113590871A (en) | Audio classification method and device and computer readable storage medium | |
| US20230178100A1 (en) | Tail point detection method, electronic device, and non-transitory computer-readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |
|
| EEER | Examination request |
Effective date: 20220608 |