Neural networks for speaker verification

Patent metadata

Field	Value
Publication number	US-11961525-B2
Application number	US-202117444384-A
Country	US
Kind code	B2
Filing date	Aug 3, 2021
Priority date	Sep 4, 2015
Publication date	Apr 16, 2024
Grant date	Apr 16, 2024

Abstract

This document generally describes systems, methods, devices, and other techniques related to speaker verification, including (i) training a neural network for a speaker verification model, (ii) enrolling users at a client device, and (iii) verifying identities of users based on characteristics of the users' voices. Some implementations include a computer-implemented method. The method can include receiving, at a computing device, data that characterizes an utterance of a user of the computing device. A speaker representation can be generated, at the computing device, for the utterance using a neural network on the computing device. The neural network can be trained based on a plurality of training samples that each: (i) include data that characterizes a first utterance and data that characterizes one or more second utterances, and (ii) are labeled as a matching speakers sample or a non-matching speakers sample.

First claim

What is claimed is:

1. A computer-implemented method when executed on data processing hardware of a user device causes the data processing hardware to perform operations comprising:
prompting a user of the user device to speak a plurality of enrollment utterances;
receiving audio signals that represent the user speaking the plurality of enrollment utterances, the audio signals representing the user speaking the plurality of enrollment utterances recorded by the user device;
generating, using a trained neural network, a reference speaker model based on the audio signals that represent the user speaking the plurality of enrollment utterances, the reference speaker model associated with the user and characterizing distinctive features of a voice of the user, the trained neural network comprising:
a long short-term memory (LSTM) layer configured to receive, as input, the audio signals representing the user speaking the plurality of enrollment utterances and generate an enrollment LSTM output; and
a fully-connected linear layer configured to receive, as input, the enrollment LSTM output and generate, as output, the reference speaker model, wherein the fully-connected linear layer comprises a last layer of the trained neural network;

storing the reference speaker model on memory hardware of the user device;
obtaining a plurality of audio frames representing a text-independent utterance different than each of the plurality of enrollment utterances;
generating, using the trained neural network, a speaker representation for the text-independent utterance, the speaker representation indicating distinctive features of a voice of a speaker of the text-independent utterance, wherein the speaker representation is generated as output from the fully-connected linear layer of the trained neural network;
determining that a similarity score between the speaker representation for the text-independent utterance and the reference speaker model stored on the memory hardware of the user device satisfies a similarity score threshold; and
based on determining that the similarity score satisfies the similarity score threshold, authenticating the speaker of the text-independent utterance as the user associated with the reference speaker model.

2. The computer-implemented method of claim 1 , wherein:
the LSTM layer is further configured to receive, as input, the plurality of audio frames representing the text-independent utterance and generate a verification LSTM output; and
the fully-connected linear layer is further configured to receive, as input, the verification LSTM output and generate, as output, the speaker representation for the text-independent utterance.

3. The computer-implemented method of claim 1 , wherein the operations further comprise, based on determining that the similarity score satisfies the similarity score threshold, updating the reference speaker model associated with the user of the user device based on the text-independent utterance.

4. The computer-implemented method of claim 1 , wherein the operations further comprise, in response to authenticating the speaker of the text-independent utterance as the user associated with the reference speaker model, transitioning operation of the user device from a low-power state to a more fully-featured state.

5. The computer-implemented method of claim 1 , wherein the operations further comprise, in response to authenticating the speaker of the text-independent utterance as the user associated with the reference speaker model:
processing one or more terms in the text-independent utterance; and
performing an action based on the one or more terms in the text-independent utterance.

6. The computer-implemented method of claim 1 , wherein the text-independent utterance comprises a non-predefined phrase.

7. The computer-implemented method of claim 1 , wherein the similarity score between the speaker representation and the reference speaker model comprises a cosine distance between a vector of values for the speaker representation and a vector of values for the reference speaker model.

8. The computer-implemented method of claim 1 , wherein the trained neural network is stored on the memory hardware of the user device.

9. The computer-implemented method of claim 1 , wherein obtaining the plurality of audio frames representing the text-independent utterance comprises:
receiving a raw audio signal of the text-independent utterance;
segmenting the raw audio signal of the text-independent utterance into a plurality of raw audio frames, each raw audio frame comprising a respective portion of the raw audio signal; and
converting the respective portion of the raw audio signal of each raw audio raw audio frame into respective audio features characterizing a respective segment of the text-independent utterance.

10. The computer-implemented method of claim 1 , wherein the operations further comprise, prior to generating the speaker representation for the text-independent utterance using the trained neural network, receiving the trained neural network over a network from a remote computing device.

11. A system comprising:
data processing hardware of a user device; and
memory hardware of the user device and in communication with the data processing hardware, the memory hardware storing instructions that when executed by the data processing hardware cause the data processing hardware to perform operations comprising:
prompting a user of the user device to speak a plurality of enrollment utterances;
receiving audio signals that represent the user speaking the plurality of enrollment utterances, the audio signals representing the user speaking the plurality of enrollment utterances recorded by the user device;
generating, using a trained neural network, a reference speaker model based on the audio signals that represent the user speaking the plurality of enrollment utterances, the reference speaker model associated with the user and characterizing distinctive features of a voice of the user, the trained neural network comprising:
a long short-term memory (LSTM) layer configured to receive, as input, the audio signals representing the user speaking the plurality of enrollment utterances and generate an enrollment LSTM output; and
a fully-connected linear layer configured to receive, as input, the enrollment LSTM output and generate, as output, the reference speaker model, wherein the fully-connected linear layer comprises a last layer of the trained neural network;

storing the reference speaker model on memory hardware of the user device;
obtaining a plurality of audio frames representing a text-independent utterance comprising a different utterance from each of the plurality of enrollment utterances;
generating, using the trained neural network, a speaker representation for the text-independent utterance, the speaker representation indicating distinctive features of a voice of a speaker of the text-independent utterance, wherein the speaker representation is generated as output from the fully-connected linear layer of the trained neural network;
determining that a similarity score between the speaker representation for the text-independent utterance and the reference speaker model stored on the memory hardware of the user device satisfies a similarity score threshold; and
based on determining that the similarity score satisfies the similarity score threshold, authenticating the speaker of the text-independent utterance as the user associated with the reference speaker model.

12. The system of claim 11 , wherein:
the LSTM layer is further configured to receive, as input, the plurality of audio frames representing the text-independ

Assignees

• Google Llc

Inventors

• Heigold Georg
• Bengio Samuel
• Lopez Moreno Ignacio

Classifications

• G10L17/18 Primary

  Artificial neural networks; Connectionist approaches · CPC title

• G10L17/02 Primary

  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 title

• G10L17/04

  Training, enrolment or model building · CPC title

• G07C9/37 Primary

  using biometric data, e.g. fingerprints, iris scans or voice recognition · CPC title