Monitoring and activating speech process in response to a trigger phrase

• US 10,102,853 B2
• Filed: 12/17/2014
• Issued: 10/16/2018
• Est. Priority Date: 12/18/2013
• Status: Active Grant

• Current Assignee: Cirrus Logic Incorporated
• Original Assignee: Cirrus Logic Incorporated
• Inventors: Page, Michael
• Primary Examiner(s): Pullias, Jesse
• Application Number:
  US15/105,755
• Publication Number:
  US 20160379635A1
• Time in Patent Office: 1,399 Days
• Field of Search: 704231-257, 704270-275
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/32 Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command
• CPC Class Codes: G10L 21/0208 Noise filtering

Abstract

A method of processing received data representing speech comprises monitoring the received data to detect the presence of data representing a first portion of a trigger phrase in said received data. On detection of the data representing the first portion of the trigger phrase, a control signal is sent to activate a speech processing block. The received data is monitored to detect the presence of data representing a second portion of the trigger phrase in said received data. If the control signal to activate the speech processing block has previously been sent, then, on detection of the data representing the second portion of the trigger phrase, the activation of the speech processing block is maintained.

Claims

1. A method of processing received data representing speech comprising the steps of:

monitoring the received data to detect the presence of data representing a first portion of a trigger phrase in said received data;

sending, on detection of said data representing the first portion of the trigger phrase, a control signal to activate a speech processing block,

monitoring the received data to detect the presence of data representing a second portion of the trigger phrase in said received data, and further comprising, if said control signal to activate the speech processing block has previously been sent;

if the data representing the second portion of the trigger phrase is not detected, sending a deactivation command to deactivate the speech processing block; and

if said data representing the second portion of the trigger phrase is detected, maintaining the activation of said speech processing block;

after detecting the data representing the first portion of the trigger phrase;

supplying a part of the received data to an adaptive speech enhancement block, and

training the speech enhancement block to derive adapted parameters for the speech enhancement block; and

if the data representing the first portion of the trigger phrase is detected;

supplying at least a part of the received data to the speech enhancement block, operating with the adapted parameters, and

outputting enhanced data from the speech enhancement block.

2. A method as claimed in claim 1 wherein the step of monitoring the received data to detect the presence of data representing the first portion of the trigger phrase and the step of monitoring the received data to detect the presence of data representing the second portion of the trigger phrase occur in parallel.

3. A method as claimed in claim 1 wherein the step of monitoring the received data for the presence of data representing the second portion of the trigger phrase occurs as a result of detecting data representing the first portion of the trigger phrase.

4. A method as claimed in claim 1 wherein, if the data representing the second portion of the trigger phrase is detected, the step of maintaining the activation of the speech processing block comprises sending a confirmation signal to the speech processing block.

5. A method as claimed in claim 1, wherein, if the data representing the second portion of the trigger phrase is detected, the step of maintaining the activation of the speech processing block comprises not sending the deactivation command.

6. A method as claimed in claim 1 further comprising:

if the data representing the second portion of the trigger phrase is detected, sending speech data to the speech processing block.

7. A method as claimed in claim 6 wherein the speech data is the received data.

8. A method as claimed in claim 1 further comprising:

if the data representing the first portion of the trigger phrase is detected, sending speech data to the speech processing block.

9. A method as claimed in claim 8, wherein the speech data is the received data.

10. A method as claimed in claim 1 further comprising:

if the data representing the second portion of the trigger phrase is detected, supplying at least a part of the received data to the speech enhancement block, operating with the adapted parameters, and

outputting enhanced data from the speech enhancement block.

11. A method as claimed in claim 10 further comprising,

storing the received speech data, wherein the at least a part of the received data supplied to the speech enhancement block is the stored data starting from the beginning of the data representing the first portion of the trigger phrase.

12. A method as claimed in claim 10 further comprising,

storing the received speech data, wherein

the at least a part of the received data supplied to the speech enhancement block is the stored data starting from the end of the data representing the first portion of the trigger phrase.

13. A method as claimed in claim 10 further comprising,

storing the received data, wherein

the at least a part of the received data supplied to the speech enhancement block is the stored data starting from the end of the data representing the second portion of the trigger phrase.

14. A method as claimed in claim 10 wherein the at least a part of the received data supplied to the speech enhancement block, is the data received immediately after said detection.

15. A method as claimed in claim 10 wherein the speech data is the enhanced data outputted from the adaptive speech enhancement block.

16. A method as claimed in claim 1 further comprising, storing the received data.

17. A method as claimed in claim 1 wherein the first trigger phrase is the first part of the second trigger phrase.

18. A speech processor, comprising:

an input, for receiving data representing speech; and

a speech processing block,

wherein the speech processor is configured to perform a method of processing received data representing speech comprising the steps of;

monitoring the received data to detect the presence of data representing a first portion of a trigger phrase in said received data;

sending, on detection of said data representing the first portion of the trigger phrase, a control signal to activate the speech processing block, and

monitoring the received data to detect the presence of data representing a second portion of the trigger phrase in said received data, and further comprising;

if said control signal to activate the speech processing block has previously been sent;

if the data representing the second portion of the trigger phrase is not detected, sending a deactivation command to deactivate the speech processing block; and

if said data representing the second portion of the trigger phrase is detected, maintaining the activation of said speech processing block;

after detecting the data representing the first portion of the trigger phrase;

supplying a part of the received data to an adaptive speech enhancement block, and

training the speech enhancement block to derive adapted parameters for the speech enhancement block; and

if the data representing the first portion of the trigger phrase is detected;

supplying at least a part of the received data to the speech enhancement block, operating with the adapted parameters, and

outputting enhanced data from the speech enhancement block.

19. A speech processor, comprising:

an input, for receiving data representing speech; and

an output, for connection to a speech processing block,

wherein the speech processor is configured to perform a method of processing received data representing speech comprising the steps of;

monitoring the received data to detect the presence of data representing a first portion of a trigger phrase in said received data;

sending, on detection of said data representing the first portion of the trigger phrase, a control signal to activate the speech processing block, and

monitoring the received data to detect the presence of data representing a second portion of the trigger phrase in said received data, and further comprising, if said control signal to activate the speech processing block has previously been sent;

if the data representing the second portion of the trigger phrase is not detected, sending a deactivation command to deactivate the speech processing block; and

if said data representing the second portion of the trigger phrase is detected, maintaining the activation of said speech processing block;

after detecting the data representing the first portion of the trigger phrase;

supplying a part of the received data to an adaptive speech enhancement block, and

training the speech enhancement block to derive adapted parameters for the speech enhancement block; and

if the data representing the first portion of the trigger phrase is detected;

supplying at least a part of the received data to the speech enhancement block, operating with the adapted parameters, and

outputting enhanced data from the speech enhancement block.

20. A mobile device, comprising a speech processor, wherein the speech processor is configured to perform a method of processing received data representing speech comprising the steps of:

monitoring the received data to detect the presence of data representing a first portion of a trigger phrase in said received data;

sending, on detection of said data representing the first portion of the trigger phrase, a control signal to activate a speech processing block, and

monitoring the received data to detect the presence of data representing a second portion of the trigger phrase in said received data, and further comprising;

if said control signal to activate the speech processing block has previously been sent;

if the data representing the second portion of the trigger phrase is not detected, sending a deactivation command to deactivate the speech processing block; and

if said data representing the second portion of the trigger phrase is detected, maintaining the activation of said speech processing block;

after detecting the data representing the first portion of the trigger phrase;

supplying a part of the received data to an adaptive speech enhancement block, and

training the speech enhancement block to derive adapted parameters for the speech enhancement block; and

if the data representing the first portion of the trigger phrase is detected;

supplying at least a part of the received data to the speech enhancement block, operating with the adapted parameters, and

outputting enhanced data from the speech enhancement block.

21. An article of manufacture comprising:

a non-transitory computer-readable medium; and

computer-executable instructions carried on the computer readable medium, the instructions readable by a processor, the instructions, when read and executed, for causing the processor to;

monitor the received data to detect the presence of data representing a first portion of a trigger phrase in said received data;

send, on detection of said data representing the first portion of the trigger phrase, a control signal to activate a speech processing block, and

monitor the received data to detect the presence of data representing a second portion of the trigger phrase in said received data, and if said control signal to activate the speech processing block has previously been sent;

if the data representing the second portion of the trigger phrase is not detected, send a deactivation command to deactivate the speech processing block; and

if said data representing the second portion of the trigger phrase is detected, maintain the activation of said speech processing block;

after detecting the data representing the first portion of the trigger phrase;

supply a part of the received data to an adaptive speech enhancement block, and

train the speech enhancement block to derive adapted parameters for the speech enhancement block; and

if the data representing the first portion of the trigger phrase is detected;

supply at least a part of the received data to the speech enhancement block, operating with the adapted parameters, and output enhanced data from the speech enhancement block.

22. A method of processing speech data comprising the steps of:

activating a speech processing block, on detecting data representing a first portion of a trigger phrase in said speech data;

maintaining said activation of said speech processing block, on subsequently detecting data representing a second portion of said trigger phrase;

de-activating said speech processing block, on subsequently detecting the absence of data representing said second portion of said trigger phrase;

after detecting the data representing the first portion of the trigger phrase;

supplying a part of the received data to an adaptive speech enhancement block, and

training the speech enhancement block to derive adapted parameters for the speech enhancement block; and

if the data representing the first portion of the trigger phrase is detected;

supplying at least a part of the received data to the speech enhancement block, operating with the adapted parameters, and

outputting enhanced data from the speech enhancement block.