CN-115910072-B - Voiceprint recognition method and device, wearable equipment and computer readable medium

Abstract

The application discloses a voiceprint recognition method, a voiceprint recognition device, wearable equipment and a computer readable medium, and relates to the technical field of voiceprint recognition; if the first voiceprint feature cannot be identified through voiceprint, acquiring a first gesture coordinate based on gesture information of the wearable equipment acquired by the position acquisition device at a designated moment, estimating a second voiceprint feature from a voiceprint feature image acquired in advance based on the first gesture coordinate, and executing voiceprint identification operation on the second voiceprint feature. Therefore, when the user cannot pass voiceprint recognition due to the deformation of the voice caused by the change of the gesture, the second voiceprint feature is determined by acquiring the first gesture coordinates according to the voiceprint feature image acquired in advance, and voiceprint recognition is performed on the second voiceprint feature, so that the voiceprint recognition passing rate when the user wears the wearable device can be improved.

Inventors

• Huang Runqian
• ZHANG WEIBIN
• CHEN DONGPENG

Assignees

• 深圳市声扬科技有限公司

Dates

Publication Date

20260508

Application Date

20221201

Claims (11)

1. 1. A voiceprint recognition method, characterized by being applied to a wearable device, the wearable device comprising an audio acquisition device and a position acquisition device, the method comprising: When the wearable device is worn on a user and used for voiceprint recognition scenes, acquiring a first voiceprint feature based on voice information of the user acquired by the audio acquisition device at a designated moment; performing a voiceprint recognition operation on the first voiceprint feature; If the first voiceprint feature cannot be identified through voiceprint, acquiring a first gesture coordinate based on gesture information of the wearable equipment acquired by the position acquisition device at the appointed moment; Estimating a second voiceprint feature from a voiceprint feature image acquired in advance based on the first gesture coordinate, wherein the voiceprint feature image is used for representing the corresponding relation between the gesture coordinate and the voiceprint feature; and executing voiceprint recognition operation on the second voiceprint feature.
2. 2. The method of claim 1, wherein before the acquiring the first voiceprint feature based on the voice information of the user acquired by the audio acquisition device at the specified time, further comprises: and training and acquiring the voiceprint feature portrait based on the voice information of the user acquired in advance by the audio acquisition device and the gesture information of the wearable equipment acquired in advance by the position acquisition device.
3. 3. A method according to claim 2, wherein training to obtain a voiceprint feature image based on the voice information of the user previously acquired by the audio acquisition device and the posture information of the wearable device previously acquired by the position acquisition device comprises: Acquiring reference voiceprint features based on voice information of the user acquired by the audio acquisition device under the condition that the user is located at a first posture position, wherein the first posture position is a position where the user is in a standing posture and the wearable equipment is located at an initial angle; acquiring gesture voiceprint features based on the voice information of the user acquired by the audio acquisition device under the condition that the user is located at a second gesture position, wherein the second gesture position is an arbitrary position different from the first gesture position; acquiring a first voiceprint deviation based on the reference voiceprint feature and the gesture voiceprint feature; acquiring a reference gesture coordinate based on gesture information of the wearable equipment acquired by the position acquisition device under the condition that the user is positioned at the first gesture position; Acquiring a second gesture coordinate based on gesture information of the wearable device acquired by the position acquisition device under the condition that the user is located at the second gesture position, wherein the second gesture coordinate is in a reference coordinate system determined by taking the reference gesture coordinate as a benchmark; and training to obtain the voiceprint feature image based on the first voiceprint deviation and the second gesture coordinate, wherein the voiceprint feature image is used for representing the corresponding relation between the gesture coordinate of the reference coordinate system and the voiceprint feature.
4. 4. A method according to claim 3, wherein the value of each of the coordinate parameters in the second pose coordinates is different from the value of the corresponding coordinate parameter in the reference pose coordinates.
5. 5. A method according to claim 3, further comprising, after said performing a voiceprint recognition operation on said second voiceprint feature: and updating the voiceprint feature image based on the second voiceprint feature.
6. 6. The method of claim 5, wherein updating the voiceprint feature representation based on the second voiceprint feature comprises: Acquiring a second voiceprint deviation based on the second voiceprint feature and the first voiceprint feature; and updating the voiceprint feature image based on the second voiceprint deviation and the first voiceprint deviation.
7. 7. The method of claim 6, wherein the first voiceprint deviation satisfies a gaussian distribution, the updating the voiceprint feature representation based on the second voiceprint deviation and the first voiceprint deviation, further comprising: Comparing the standard deviation of the Gaussian distribution of the second voiceprint deviation and the first voiceprint deviation, and updating the mean value of the Gaussian distribution of the first voiceprint deviation when the second voiceprint deviation is larger than the standard deviation; And updating the voiceprint feature image based on the updated first voiceprint deviation and the second gesture coordinate.
8. 8. A method according to claim 1, wherein said estimating a second voiceprint feature from a pre-acquired voiceprint feature image based on said first pose coordinates comprises: acquiring a reference gesture coordinate based on gesture information of the wearable device acquired by the position acquisition device under the condition that the user is located at a first gesture position, wherein the first gesture position is a position where the user is in a standing gesture and the wearable device is located at an initial angle; Acquiring a reference voiceprint feature based on the voice information of the user acquired by the audio acquisition device under the condition that the user is positioned at the first gesture position; Interpolating to obtain a third gesture coordinate based on the reference gesture coordinate and the first gesture coordinate; and acquiring the second voiceprint feature from the voiceprint feature distribution portrait based on the third gesture coordinate.
9. 9. Voiceprint recognition device, characterized in that is applied to wearable equipment, wearable equipment includes audio acquisition device and position acquisition device, and the device includes: The voiceprint feature acquisition unit is used for acquiring voiceprint features based on the voice information of the user acquired by the audio acquisition device; The voiceprint recognition unit is used for carrying out voiceprint recognition on the voiceprint features; The gesture coordinate acquisition unit is used for acquiring gesture coordinates based on the gesture information of the wearable equipment acquired by the position acquisition device when the voiceprint characteristics acquired by the voiceprint characteristic acquisition unit do not pass voiceprint recognition; And the voiceprint feature analysis unit is used for estimating voiceprint features from the voiceprint feature images acquired in advance based on the gesture coordinates.
10. 10. A wearable device, comprising: One or more processors; A memory; an audio acquisition device and a position acquisition device; one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform the method of any of claims 1-8.
11. 11. A computer readable medium, characterized in that the computer readable medium stores a program code executable by a processor, which program code, when executed by the processor, causes the processor to perform the method of any of claims 1-8.

Description

Voiceprint recognition method and device, wearable equipment and computer readable medium Technical Field The present application relates to the technical field of voiceprint recognition, and in particular, to a voiceprint recognition method, a voiceprint recognition device, a wearable device, and a computer readable medium. Background At present, in order to meet the use requirements of different scenes and improve the user interaction experience, a voiceprint recognition module is often carried on the wearable device, but if a wearer has some posture changes in the use process, such as squatting, leaning back, bending over and the like, the sounding organ of the wearer is pressed, the emitted sound is also greatly changed, and therefore the passing rate of voiceprint recognition is reduced. Disclosure of Invention The application provides a voiceprint recognition method, a voiceprint recognition device, wearable equipment and a computer readable medium, so as to improve the defects. In a first aspect, an embodiment of the application provides a voiceprint recognition method, which is applied to a wearable device, wherein the wearable device comprises an audio acquisition device and a position acquisition device, and the method comprises the steps of acquiring a first voiceprint feature based on voice information of a user acquired by the audio acquisition device at a designated moment when the wearable device is worn on the user and is used in a voiceprint recognition scene, executing voiceprint recognition operation on the first voiceprint feature, acquiring a first gesture coordinate based on gesture information of the wearable device acquired by the position acquisition device at the designated moment if the first voiceprint feature cannot pass voiceprint recognition, estimating a second voiceprint feature based on the first gesture coordinate, and executing the voiceprint recognition operation on the second voiceprint feature based on a pre-acquired voiceprint feature image, wherein the voiceprint feature image is used for representing a corresponding relation between the gesture coordinate and the voiceprint feature. In a second aspect, the embodiment of the application also provides a voiceprint recognition device which is applied to the wearable equipment, wherein the wearable equipment comprises an audio acquisition device and a position acquisition device, and the device comprises a voiceprint feature acquisition unit, a voiceprint recognition unit, a gesture coordinate acquisition unit and a voiceprint feature analysis unit. The voice-print recognition system comprises a voice-print feature acquisition unit, a voice-print recognition unit, a gesture coordinate acquisition unit and a voice-print feature analysis unit, wherein the voice-print feature acquisition unit is used for acquiring voice-print features based on voice information of a user acquired by the voice-print acquisition device, the voice-print recognition unit is used for carrying out voice-print recognition on the voice-print features, the gesture coordinate acquisition unit is used for acquiring gesture coordinates based on gesture information of the wearable device acquired by the position acquisition device when the voice-print features acquired by the voice-print feature acquisition unit do not pass voice-print recognition, and the voice-print feature analysis unit is used for analyzing and acquiring the voice-print features based on the gesture coordinates. In a third aspect, embodiments of the present application also provide a wearable device comprising one or more processors, memory, audio acquisition means and location acquisition means, one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform a method as described above. In a fourth aspect, embodiments of the present application also provide a computer readable medium storing program code executable by a processor, the program code when executed by the processor causing the processor to perform the above method. The application provides a voiceprint recognition method, a voiceprint recognition device, a wearable device and a computer readable medium, in the case that the wearable device is worn on a user and used for voiceprint recognition scenes, and acquiring a first voiceprint feature based on the voice information of the user acquired by the audio acquisition device at the appointed moment, and executing voiceprint recognition operation on the first voiceprint feature. Then, if the first voiceprint feature cannot pass voiceprint recognition, based on the gesture information of the wearable device acquired by the position acquisition device at the appointed moment, acquiring a first gesture coordinate, and estimating a second voiceprint feature from a voiceprint feature image acquired in advance, wherein the voiceprint feature i