# US 2020/0349955 A1 · System and method for speaker recognition on mobile devices

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0349955 A1
- Document ID: US20200349955A1
- Status: Published application
- Kind: A1
- Application number: 16/920,137
- Filed: 2020-07-02
- Published: 2020-11-05
- Earliest claimed: 2012-02-24
- Assignees: Cirrus Logic International Semiconductor Ltd.
- Applicants: Cirrus Logic International Semiconductor Ltd.
- Inventors: Marta Garcia GOMAR
- Web page: https://patentgrep.com/patent/US20200349955A1

## CPC

- G10L 17/00: Speaker identification or verification techniques
- G10L 17/22: Interactive procedures; Man-machine interfaces
- H04L 63/0861: {using biometrical features, e.g. fingerprint, retina-scan (cryptographic mechanisms or cryptographic arrangements for entity authentication using biological data H04L9/3231)}
- H04W 12/06: Authentication
- G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)
- H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72)

## Abstract

A speaker recognition system for authenticating a mobile device user includes an enrollment and learning software module, a voice biometric authentication software module, and a secure software application. Upon request by a user of the mobile device, the enrollment and learning software module displays text prompts to the user, receives speech utterances from the user, and produces a voice biometric print. The enrollment and training software module determines when a voice biometric print has met at least a quality threshold before storing it on the mobile device. The secure software application prompts a user requiring authentication to repeat an utterance based at least on an attribute of a selected voice biometric print, receives a corresponding utterance, requests the voice biometric authentication software module to verify the identity of the second user using the utterance, and, if the user is authenticated, imports the voice biometric print.

## Claims

### Claim at position 1

1-17. (canceled)

### Claim 18

A method of speaker authentication, comprising:
in a mobile device, obtaining an audio signal representing speech;
in the mobile device, extracting features of the audio signal;
transmitting the extracted features from the mobile device to a server;
in the server, performing a speaker authentication process using the extracted features; and
transmitting a result of the speaker authentication process from the server to the mobile device.

### Claim 19 (depends on 18)

A method according to claim 18, further comprising:
in the mobile device, after extracting features of the audio signal, performing first normalizations.

### Claim 20 (depends on 19)

A method according to claim 19, further comprising:
in the server, after receiving the extracted features of the audio signal from the mobile device, performing second normalizations, wherein the second normalizations are different from the first normalizations performed in the mobile device.

### Claim 21 (depends on 18)

A method according to claim 18, further comprising:
in the server, after receiving features of the audio signal, performing normalizations.

### Claim 22 (depends on 18)

A method according to claim 18, wherein the speaker authentication process comprises a process for determining whether the audio signal represents the speech of an enrolled user of the mobile device.

### Claim 23 (depends on 22)

A method according to claim 22, wherein the result of the speaker authentication process is either a first value indicating that the audio signal represents the speech of an enrolled user of the mobile device or a second value indicating that the audio signal does not represent the speech of an enrolled user of the mobile device.

### Claim 24 (depends on 18)

A method according to claim 18, wherein the speaker authentication process comprises a process for identifying a person whose speech is represented by the audio signal.

### Claim 25 (depends on 24)

A method according to claim 24, wherein the result of the speaker authentication process is either an identification of the speaker or an indication that the speaker cannot be identified.

### Claim 26 (depends on 18)

A method according to claim 18, further comprising, after extracting the features of the audio signal, forming an encrypted feature vector, wherein transmitting the extracted features from the mobile device to a server comprises transmitting the encrypted feature vector from the mobile device to the server.

### Claim 27 (depends on 26)

A method according to claim 26, further comprising, after forming the encrypted feature vector, compressing the encrypted feature vector, wherein transmitting the extracted features from the mobile device to a server comprises transmitting the compressed encrypted feature vector from the mobile device to the server.

### Claim 28 (depends on 18)

A method according to claim 18, further comprising, after extracting the features of the audio signal, forming an feature vector, and further comprising compressing the feature vector, wherein transmitting the extracted features from the mobile device to a server comprises transmitting the compressed feature vector from the mobile device to the server.

### Claim 29

A mobile device, comprising:
an input for obtaining an audio signal representing speech; and
a processor configured for extracting features of the audio signal,
wherein the mobile device is configured for transmitting the extracted features from the mobile device to a server, and
wherein the mobile device is further configured for receiving from the server a result of a speaker authentication process performed using the extracted features.

### Claim 30 (depends on 29)

A mobile device according to claim 29, wherein the processor is further configured for, after extracting features of the audio signal, performing first normalizations.

### Claim 31 (depends on 29)

A mobile device according to claim 29, wherein the mobile device is further configured for, after extracting the features of the audio signal, forming an encrypted feature vector.

### Claim 32 (depends on 31)

A mobile device according to claim 31, wherein the mobile device is further configured for transmitting the encrypted feature vector to the server.

### Claim 33 (depends on 31)

A mobile device according to claim 31, wherein the mobile device is further configured for, after forming the encrypted feature vector, compressing the encrypted feature vector, and transmitting the compressed encrypted feature vector to the server.

### Claim 34

A server, comprising:
a receiver, configured for receiving from a mobile device features extracted by the mobile device from an audio signal representing speech;
a processor, configured for performing a speaker authentication process using the received extracted features; and
a transmitter, for transmitting a result of the speaker authentication process from the server to the mobile device.

### Claim 35 (depends on 34)

A server according to claim 34, wherein the processor is further configured for performing normalizations on the received extracted features.

### Claim 36 (depends on 34)

A server according to claim 34, wherein the speaker authentication process comprises a process for determining whether the audio signal represents the speech of an enrolled user of the mobile device.

### Claim 37 (depends on 36)

A server according to claim 36, wherein the result of the speaker authentication process is either a first value indicating that the audio signal represents the speech of an enrolled user of the mobile device or a second value indicating that the audio signal does not represent the speech of an enrolled user of the mobile device.
