Low power detection of a voice control activation phrase

• US 9,767,798 B2
• Filed: 03/20/2015
• Issued: 09/19/2017
• Est. Priority Date: 02/27/2015
• Status: Active Grant

• Current Assignee: Imagination Technologies Limited (Imagination Technologies Group Ltd.)
• Original Assignee: Imagination Technologies Limited (Imagination Technologies Group Ltd.)
• Inventors: Kerr, Benjamin James
• Primary Examiner(s): Jerez Lora, William A
• Application Number:
  US14/663,583
• Publication Number:
  US 20160253997A1
• Time in Patent Office: 914 Days
• Field of Search: 381 56, 381 58, 381110
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/28 Constructional details of speech recognition systems
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command

Abstract

Methods of low power detection of an activation phrase are described. A microphone system comprises dedicated hardware logic for detecting a pre-defined activation phrase in an audio stream received via a microphone. If the pre-defined activation phrase is detected, the hardware logic sends a trigger signal to activate a module, such as a main speech detection module, which is external to the microphone system and which may be in a low power standby state.

Claims

1. A microphone system comprising a microphone and hardware logic configured to compare an audio stream received via the microphone with a pre-defined phrase and in response to partially detecting the pre-defined phrase in a first portion of the audio stream, to:

determine if a previous portion of the audio stream comprising a partially detected pre-defined phrase was stored;

in response to determining that the previous portion of the audio stream comprising a partially detected pre-defined phrase was not stored, to store the first portion of the audio stream;

in response to determining that the previous portion of the audio stream comprising a partially detected pre-defined phrase was stored, to compare the first portion of the audio stream and the stored previous portion of the audio stream; and

in response to detecting a match between the first portion of the audio stream and the stored previous portion of the audio stream, to send a trigger signal to activate a module external to the microphone system.

2. A microphone system according to claim 1, further comprising an audio buffer arranged to store a second portion of the audio stream received via the microphone after a first portion of the audio stream comprising the pre-defined phrase.

3. A microphone system according to claim 1, wherein the hardware logic is configured to divide a portion of detected audio into one or more component parts and to compare the component parts to one or more stored component parts of the pre-defined phrase.

4. A microphone system according to claim 1, further comprising an amplifier configured to amplify the audio stream received via the microphone according to a gain parameter and to output the amplified audio stream to the hardware logic which is configured to detect the pre-defined phrase, and wherein the hardware logic is further configured, in response to detecting the pre-defined phrase, to send a signal to the amplifier to increase a value of the gain parameter.

5. A microphone system according to claim 4, wherein the hardware logic is further configured to send a signal to the amplifier to reduce the value of the gain parameter in response to the module external to the microphone system entering a low power mode.

6. A microphone system according to claim 1, further comprising an analogue to digital converter configured to convert the audio stream received via the microphone into a digital audio stream by sampling the received audio stream at a sample rate and a sample resolution and to output the digital audio stream to the hardware logic which is configured to detect the pre-defined phrase and wherein the hardware logic is further configured, in response to detecting the pre-defined phrase, to send a signal to increase the sample rate and/or sample resolution.

7. A microphone system according to claim 6, wherein the hardware logic is further configured to send a signal to reduce the sample rate and/or sample resolution in response to the module external to the microphone system entering a low power mode.

8. A microphone system according to claim 1, further comprising a memory arranged to store the pre-defined phrase, wherein the pre-defined phrase is an activation phrase for activating a module external to the microphone system.

9. A microphone system according to claim 8, wherein the hardware logic is further configured, in response to detecting the pre-defined phrase, to update the stored pre-defined phrase using the detected pre-defined phrase.

10. A microphone system according to claim 1, wherein the hardware logic is further configured, in response to partially detecting the pre-defined phrase, to send a portion of the audio stream comprising the partially detected pre-defined phrase to the module external to the microphone system.

11. A microphone system according to claim 1, wherein the pre-defined phrase is a sample of a contextual sound.

12. A microphone system according to claim 1, wherein the pre-defined phrase comprises one or more spoken words.

13. A microphone system according to claim 12, wherein the pre-defined phrase is an activation phrase for a voice controlled computing device.

14. A microphone system according to claim 13, wherein the module external to the microphone system is a main speech detection module and the main speech detection module is implemented by a processor in the voice controlled computing device.

15. A microphone system according to claim 12, wherein the microphone system comprises two modes of operation and wherein

in a first mode of operation, the hardware logic is configured to detect a pre-defined phrase in an audio stream received via the microphone by dividing a portion of detected audio into one or more component parts and comparing the component parts to one or more stored component parts of the pre-defined phrase and in response to detecting the pre-defined phrase, to send a trigger signal to activate a first module external to the microphone system;

and

in a second mode of operation, the hardware logic is configured to divide a portion of detected audio into one or more component parts and to provide the component parts to a second module external to the microphone system.

16. A microphone system according to claim 15, wherein the first module external to the microphone system and the second module external to the microphone system are the same and the module is a main speech detection module.

17. A microphone system according to claim 15, wherein the first module external to the microphone system is a main speech detection module and the second module external to the microphone system is a communications interface arranged to transmit the component parts to a remote phrase detection and recognition system.

18. A microphone system according to claim 1, wherein the hardware logic is configured to continuously analyse the audio stream received via the microphone to detect the pre-defined phrase.

19. A method of operating a microphone system comprising:

detecting an audio stream from a microphone in the microphone system;

comparing the detected audio stream with a pre-defined phrase stored in dedicated hardware in the microphone system; and

in response to partially detecting the pre-defined phrase in a first portion of the audio stream;

determining if a previous portion of the audio stream comprising a partially detected pre-defined phrase was stored;

in response to determining that the previous portion of the audio stream comprising a partially detected pre-defined phrase was not stored, storing the first portion of the audio stream;

in response to determining that the previous portion of the audio stream comprising a partially detected pre-defined phrase was stored, comparing the first portion of the audio stream and the stored previous portion of the audio stream; and

in response to detecting a match between the first portion of the audio stream and the stored previous portion of the audio stream, sending a trigger signal to activate a module external to the microphone system.