METHOD AND DEVICE FOR VOICEPRINT RECOGNITION

• US 20140214417A1
• Filed: 12/12/2013
• Published: 07/31/2014
• Est. Priority Date: 01/28/2013
• Status: Active Grant

• Current Assignee: Tencent Technology Company Limited (Tencent Holdings Limited)
• Original Assignee: Tencent Technology Company Limited (Tencent Holdings Limited)
• Inventors: ZHANG, Xiang, LIU, Haibo, CHEN, Bo, WANG, Eryu, LU, Li, LI, Lou, RAO, Feng, LU, Duling, YUE, Shuai
• Granted Patent: US 9,502,038 B2
• Time in Patent Office: Days
• US Class Current: 704/232
• CPC Class Codes: G10L 17/02 Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction
• CPC Class Codes: G10L 17/04 Training, enrolment or model building
• CPC Class Codes: G10L 17/08 Use of distortion metrics or a particular distance between probe pattern and reference templates
• CPC Class Codes: G10L 17/18 Artificial neural networks; Connectionist approaches

Abstract

A method and device for voiceprint recognition, include: establishing a first-level Deep Neural Network (DNN) model based on unlabeled speech data, the unlabeled speech data containing no speaker labels and the first-level DNN model specifying a plurality of basic voiceprint features for the unlabeled speech data; obtaining a plurality of high-level voiceprint features by tuning the first-level DNN model based on labeled speech data, the labeled speech data containing speech samples with respective speaker labels, and the tuning producing a second-level DNN model specifying the plurality of high-level voiceprint features; based on the second-level DNN model, registering a respective high-level voiceprint feature sequence for a user based on a registration speech sample received from the user; and performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user.

Claims

1. A method of voiceprint recognition, comprising:

at a device having one or more processors and memory;

establishing a first-level Deep Neural Network (DNN) model based on unlabeled speech data, the unlabeled speech data containing no speaker labels and the first-level DNN model specifying a plurality of basic voiceprint features for the unlabeled speech data;

obtaining a plurality of high-level voiceprint features by tuning the first-level DNN model based on labeled speech data, the labeled speech data containing speech samples with respective speaker labels, and the tuning producing a second-level DNN model specifying the plurality of high-level voiceprint features;

based on the second-level DNN model, registering a respective high-level voiceprint feature sequence for a user based on a registration speech sample received from the user; and

performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user.

2. The method of claim 1, wherein tuning the first-level DNN model based on the labeled speech data to produce the second-level DNN model further comprises:

imposing at least two constraints during adjustment of the first-level DNN based on the labeled speech data, including;

(1) distances between characteristic voiceprints generated from speech samples of different speakers increase with training, and (2) distances between characteristic voice prints generated from speech samples of same speakers decrease with training.

3. The method of claim 1, wherein tuning the first-level DNN model based on the labeled speech data further comprises:

applying sparse coding limit rules to train the first-level DNN model based on the labeled speech data.

4. The method of claim 1, wherein tuning the first-level DNN model based on the labeled speech data further comprises:

applying maximum cross entropy rules to train the first-level DNN model based on the labeled speech data.

5. The method of claim 1, wherein, based on the second-level DNN model, registering the respective high-level voiceprint feature sequence for the user based on the registration speech sample received from the user, further comprises:

obtaining a respective basic voiceprint feature sequence for the user from the registration speech sample based on the first-level DNN model; and

providing the respective basic voiceprint feature sequence as input to the second-level DNN model to obtain the respective high-level voiceprint feature sequence for the user.

6. The method of claim 1, wherein performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user, further comprise:

receiving a test speech sample;

obtaining a respective high-level voiceprint feature sequence from the test speech sample using the first-level DNN model and the second-level DNN model in sequence; and

comparing the respective high-level voiceprint feature sequence obtained from the test speech sample to the respective high-level voiceprint feature sequence registered to the user.

7. The method of claim 6, wherein performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user, further comprise:

verifying the user'"'"'s identity based on a Euclidean distance between respective Gaussian models of the respective high-level voiceprint feature sequence obtained from the test speech sample and the respective high-level voiceprint feature sequence registered to the user.

8. A voiceprint recognition system, comprising:

one or more processors; and

memory having instructions stored thereon, the instructions, when executed by the one or more processors, cause the processors to perform operations comprising;

establishing a first-level Deep Neural Network (DNN) model based on unlabeled speech data, the unlabeled speech data containing no speaker labels and the first-level DNN model specifying a plurality of basic voiceprint features for the unlabeled speech data;

obtaining a plurality of high-level voiceprint features by tuning the first-level DNN model based on labeled speech data, the labeled speech data containing speech samples with respective speaker labels, and the tuning producing a second-level DNN model specifying the plurality of high-level voiceprint features;

based on the second-level DNN model, registering a respective high-level voiceprint feature sequence for a user based on a registration speech sample received from the user; and

performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user.

9. The system of claim 8, wherein tuning the first-level DNN model based on the labeled speech data to produce the second-level DNN model further comprises:

imposing at least two constraints during adjustment of the first-level DNN based on the labeled speech data, including;

(1) distances between characteristic voiceprints generated from speech samples of different speakers increase with training, and (2) distances between characteristic voice prints generated from speech samples of same speakers decrease with training.

10. The system of claim 8, wherein tuning the first-level DNN model based on the labeled speech data further comprises:

applying sparse coding limit rules to train the first-level DNN model based on the labeled speech data.

11. The system of claim 8, wherein tuning the first-level DNN model based on the labeled speech data further comprises:

applying maximum cross entropy rules to train the first-level DNN model based on the labeled speech data.

12. The system of claim 8, wherein, based on the second-level DNN model, registering the respective high-level voiceprint feature sequence for the user based on the registration speech sample received from the user, further comprises:

obtaining a respective basic voiceprint feature sequence for the user from the registration speech sample based on the first-level DNN model; and

providing the respective basic voiceprint feature sequence as input to the second-level DNN model to obtain the respective high-level voiceprint feature sequence for the user.

13. The system of claim 8, wherein performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user, further comprise:

receiving a test speech sample;

obtaining a respective high-level voiceprint feature sequence from the test speech sample using the first-level DNN model and the second-level DNN model in sequence; and

comparing the respective high-level voiceprint feature sequence obtained from the test speech sample to the respective high-level voiceprint feature sequence registered to the user.

14. The system of claim 13, wherein performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user, further comprise:

verifying the user'"'"'s identity based on a Euclidean distance between respective Gaussian models of the respective high-level voiceprint feature sequence obtained from the test speech sample and the respective high-level voiceprint feature sequence registered to the user.

15. A non-transitory computer-readable medium having instructions stored thereon, the instructions, when executed by one or more processors, cause the processors to perform operations comprising:

establishing a first-level Deep Neural Network (DNN) model based on unlabeled speech data, the unlabeled speech data containing no speaker labels and the first-level DNN model specifying a plurality of basic voiceprint features for the unlabeled speech data;

obtaining a plurality of high-level voiceprint features by tuning the first-level DNN model based on labeled speech data, the labeled speech data containing speech samples with respective speaker labels, and the tuning producing a second-level DNN model specifying the plurality of high-level voiceprint features;

based on the second-level DNN model, registering a respective high-level voiceprint feature sequence for a user based on a registration speech sample received from the user; and

performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user.

16. The computer-readable medium of claim 15, wherein tuning the first-level DNN model based on the labeled speech data to produce the second-level DNN model further comprises:

imposing at least two constraints during adjustment of the first-level DNN based on the labeled speech data, including;

(1) distances between characteristic voiceprints generated from speech samples of different speakers increase with training, and (2) distances between characteristic voice prints generated from speech samples of same speakers decrease with training.

17. The computer-readable medium of claim 15, wherein tuning the first-level DNN model based on the labeled speech data further comprises:

applying sparse coding limit rules to train the first-level DNN model based on the labeled speech data.

18. The computer-readable medium of claim 15, wherein tuning the first-level DNN model based on the labeled speech data further comprises:

applying maximum cross entropy rules to train the first-level DNN model based on the labeled speech data.

19. The computer-readable medium of claim 15, wherein, based on the second-level DNN model, registering the respective high-level voiceprint feature sequence for the user based on the registration speech sample received from the user, further comprises:

obtaining a respective basic voiceprint feature sequence for the user from the registration speech sample based on the first-level DNN model; and

providing the respective basic voiceprint feature sequence as input to the second-level DNN model to obtain the respective high-level voiceprint feature sequence for the user.

20. The computer-readable medium of claim 15, wherein performing speaker verification for the user based on the respective high-level voiceprint feature sequence registered for the user, further comprise:

receiving a test speech sample;

obtaining a respective high-level voiceprint feature sequence from the test speech sample using the first-level DNN model and the second-level DNN model in sequence; and

comparing the respective high-level voiceprint feature sequence obtained from the test speech sample to the respective high-level voiceprint feature sequence registered to the user.