Promoting voice actions to hotwords

Patent metadata

Field	Value
Publication number	US-9263035-B2
Application number	US-201414221520-A
Country	US
Kind code	B2
Filing date	Mar 21, 2014
Priority date	Dec 5, 2013
Publication date	Feb 16, 2016
Grant date	Feb 16, 2016

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for designating certain voice commands as hotwords. The methods, systems, and apparatus include actions of receiving a hotword followed by a voice command. Additional actions include determining that the voice command satisfies one or more predetermined criteria associated with designating the voice command as a hotword, where a voice command that is designated as a hotword is treated as a voice input regardless of whether the voice command is preceded by another hotword. Further actions include, in response to determining that the voice command satisfies one or more predetermined criteria associated with designating the voice command as a hotword, designating the voice command as a hotword.

First claim

The invention claimed is:

1. A computer-implemented method comprising:
determining, by a device, that a voice command which includes a first term followed by a second term has been newly designated as a hotword;
determining, by the device, that an existing hotword detector is configured to detect utterances that include the first term followed by a third term that is different from the second term, wherein a hotword detector comprises a neural network that is configured to estimate a likelihood of a match between (i) acoustic feature associated with an utterance and (ii) corresponding acoustic features associated with representative utterances of a particular term, without transcribing the utterance; and
enabling, by the device, a set of hotword detectors including a first, a second, and a third hotword detector configured to detect utterances that include the first term, the second term, and the third term, respectively.

2. The method of claim 1 , wherein a voice command that is designated as a hotword is treated as a voice input regardless of whether the voice command is preceded by another hotword.

3. The method of claim 1 , comprising:
disabling the existing hotword detector that is configured to detect utterances that include the first term followed by the different, third term.

4. The method of claim 1 , comprising:
receiving an utterance including the first term followed by another term;
receiving an output, from the first hotword detector, that the utterance includes the first term; and
in response to receiving the output, determining whether the other term is the second term or the third term using the second hotword detector and the third hotword detector, respectively.

5. The method of claim 1 , comprising:
in response to determining whether the other term is the second term or the third term using the second hotword detector and the third hotword detector, respectively, determining that the other term is the second term; and
in response determining that the other term is the second term, determining that the utterance is a voice input.

6. The method of claim 1 , comprising:
in response to determining whether the other term is the second term or the third term using the second hotword detector and the third hotword detector, respectively, determining that the other term is the third term; and
in response determining that the other term is the third term, determining that the utterance is a voice input.

7. The method of claim 1 , comprising:
in response to determining whether the other term is the second term or the third term using the second hotword detector and the third hotword detector, respectively, determining that the other term is not the second term or the third term; and
in response determining that the other term is not the second term or the third term, determining that the utterance is not a voice input.

8. The method of claim 1 , comprising:
determining that a second voice command which includes the first term followed by a different, fourth term has been newly designated as a hotword;
in response to determining that the voice command which includes the first term followed by the second term has been designated as a hotword, determining that an existing hotword detector is configured to detect utterances that include the first term; and
in response to determining that an existing hotword detector is configured to detect utterances that include the first term, enabling a fourth hotword detector configured to detect utterances that include the different, fourth term.

9. The method of claim 1 , wherein a hotword comprises (i) a term that wakes the device up from a sleep state or hibernation state or (ii) a term that triggers semantic interpretation on the term or on one or more terms that follow the term.

10. A system comprising:
one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
determining, by a device, that a voice command which includes a first term followed by a second term has been newly designated as a hotword;
determining, by the device, that an existing hotword detector is configured to detect utterances that include the first term followed by a third term that is different from the second term wherein a hotword detector comprises a neural network that is configured to estimate a likelihood of a match between (i) acoustic feature associated with an utterance and (ii) corresponding acoustic features associated with representative utterances of a particular term, without transcribing the utterance; and
enabling, by the device, a set of hotword detectors including a first, a second, and a third hotword detector configured to detect utterances that include the first term, the second term, and the third term, respectively.

11. The system of claim 10 , wherein a voice command that is designated as a hotword is treated as a voice input regardless of whether the voice command is preceded by another hotword.

12. The system of claim 10 , the operations comprising:
disabling the existing hotword detector that is configured to detect utterances that include the first term followed by the different, third term.

13. The system of claim 10 , the operations comprising:
receiving an utterance including the first term followed by another term;
receiving an output, from the first hotword detector, that the utterance includes the first term; and
in response to receiving the output, determining whether the other term is the second term or the third term using the second hotword detector and the third hotword detector, respectively.

14. The system of claim 10 , the operations comprising:
in response to determining whether the other term is the second term or the third term using the second hotword detector and the third hotword detector, respectively, determining that the other term is the second term; and
in response determining that the other term is the second term, determining that the utterance is a voice input.

15. The system of claim 10 , the operations comprising:
in response to determining whether the other term is the second term or the third term using the second hotword detector and the third hotword detector, respectively, determining that the other term is the third term; and
in response determining that the other term is the third term, determining that the utterance is a voice input.

16. The system of claim 10 , the operations comprising:
in response to determining whether the other term is the second term or the third term using the second hotword detector and the third hotword detector, respectively, determining that the other term is not the second term or the third term; and
in response determining that the other term is not the second term or the third term, determining that the utterance is not a voice input.

17. The system of claim 10 , the operations comprising:
determining that a second voice command which includes the first term followed by a different, fourth term has been newly designated as a hotword;
in response to determining that the voice command which includes the first term followed by the second term has been designated as a hotword, determining that an existing hotword detector is configured to detect utterances that include the first term; and
in response to determining that an existing hotword detector is configured to detect utterances that include the first term, enabling a fourth hotword detector configured to detect utteranc

Assignees

• Google Inc

Inventors

• Sharifi Matthew

Classifications

• G10L15/22 Primary

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

• G10L2015/223

  Execution procedure of a spoken command · CPC title

• G10L15/285

  Memory allocation or algorithm optimisation to reduce hardware requirements · CPC title

• G10L15/08 Primary

  Speech classification or search · CPC title

• G06F3/167

  Audio in a user interface, e.g. using voice commands for navigating, audio feedback · CPC title