Voice command triggered speech enhancement

Patent metadata

Field	Value
Publication number	US-10319372-B2
Application number	US-201715688380-A
Country	US
Kind code	B2
Filing date	Aug 28, 2017
Priority date	Dec 18, 2013
Publication date	Jun 11, 2019
Grant date	Jun 11, 2019

Abstract

Received data representing speech is stored, and a trigger detection block detects a presence of data representing a trigger phrase in the received data. In response, a first part of the stored data representing at least a part of the trigger phrase is supplied to an adaptive speech enhancement block, which is trained on the first part of the stored data to derive adapted parameters for the speech enhancement block. A second part of the stored data, overlapping with the first part of the stored data, is supplied to the adaptive speech enhancement block operating with said adapted parameters, to form enhanced stored data. A second trigger phrase detection block detects the presence of data representing the trigger phrase in the enhanced stored data. In response, enhanced speech data are output from the speech enhancement block for further processing, such as speech recognition.

First claim

The invention claimed is:

1. A method of processing received data representing speech, comprising:
on detecting that a segment of the received data represents a trigger phrase:
selecting a part of the received data representing the trigger phrase, wherein the selected part of the received data contains speech data;
training an adaptive speech enhancement block on he selected part of the received data to derive parameters for the adaptive speech enhancement block;
passing said segment of received data through said adaptive speech enhancement block using said derived parameters to generate an enhanced segment of data; and
attempting to detect said trigger phrase in the enhanced segment of data.

2. A method as claimed in claim 1 , comprising, if unsuccessful in attempting to detect said trigger phrase in the enhanced segment of data, resetting a detector for detecting that a segment of the received data represents a trigger phrase.

3. A method as claimed in claim 1 , comprising, if successful in attempting to detect said trigger phrase in the enhanced segment of data, passing further received data through the adaptive speech enhancement block for further downstream signal processing.

4. A method as claimed in claim 1 , comprising attempting to detect said trigger phrase in the enhanced segment of data using different detection criteria from those used in detecting whether the segment of the received data represents a trigger phrase.

5. A method as claimed in claim 4 , comprising attempting to detect said trigger phrase in the enhanced segment of data using more rigorous detection criteria from those used in detecting whether the segment of the received data represents a trigger phrase.

6. A method as claimed in claim 1 , comprising:
receiving and storing data from multiple microphones;
detecting that the segment of the received data represents a trigger phrase, based on data received from a subset of said microphones;
training the adaptive speech enhancement block on at least a part of said segment of the received data received from said multiple microphones to derive said adapted parameters for the speech enhancement block; and
passing said segment of the received data received from said multiple microphones through said adaptive speech enhancement block using said derived parameters to generate the enhanced segment of data.

7. A method as claimed in claim 6 , wherein the speech enhancement block is a beamformer.

8. A method as claimed in claim 1 , comprising passing said segment of received data through said adaptive speech enhancement block and generating the enhanced segment of data at a higher rate than real time.

9. A speech processor, comprising:
an input, for receiving data representing speech; and
a speech processing block, wherein the speech processing block is configured to perform a method comprising:
on detecting that a segment of the received data represents a trigger phrase:
selecting a part of the received data representing the trigger phrase, wherein the selected part of the received data contains speech data;
training an adaptive speech enhancement block on the selected part of the received data to derive parameters for the adaptive speech enhancement block;
passing said segment of received data through said adaptive speech enhancement block using said derived parameters to generate an enhanced segment of data; and
attempting to detect said trigger phrase in the enhanced segment of data.

10. A mobile device, comprising a speech processor as claimed in claim 9 .

11. A speech processor, comprising:
an input, for receiving data representing speech; and
an output, for connection to a speech processing block, wherein the speech processing block is configured to perform a method comprising:
on detecting that a segment of the received data represents a trigger phrase:
selecting a part of the received data representing the trigger phrase, wherein the selected part of the received data contains speech data;
training an adaptive speech enhancement block on the selected part of the received data to derive parameters for the adaptive speech enhancement block;
passing said segment of received data through said adaptive speech enhancement block using said derived parameters to generate an enhanced segment of data; and
attempting to detect said trigger phrase in the enhanced segment of data.

12. A mobile device, comprising a speech processor as claimed in claim 11 .

13. A computer program product, comprising computer readable code embodied in non-transitory computer-readable media, for causing a processing device to perform a method comprising:
selecting a part of the received data representing the trigger phrase, wherein the selected part of the received data contains speech data;
training an adaptive speech enhancement block on the selected part of the received data to derive parameters for the adaptive speech enhancement block;
passing said segment of received data through said adaptive speech enhancement block using said derived parameters to generate an enhanced segment of data; and
attempting to detect said trigger phrase in the enhanced segment of data.

Assignees

• Cirrus Logic Int Semiconductor Ltd
• Cirrus Logic Inc

Inventors

• Hatfield Robert James
• Page Michael

Classifications

• G10L2021/02166

  Microphone arrays; Beamforming · CPC title

• G10L21/0208 Primary

  Noise filtering · CPC title

• G10L21/02 Primary

  Speech enhancement, e.g. noise reduction or echo cancellation (reducing echo effects in line transmission systems H04B3/20; echo suppression in hands-free telephones H04M9/08) · CPC title

• G10L17/24

  the user being prompted to utter a password or a predefined phrase · CPC title

• G10L15/22 Primary

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title