Providing pre-computed hotword models
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for obtaining, for each of multiple words or sub-words, audio data corresponding to multiple users speaking the word or sub-word; training, for each of the multiple words or sub-words, a pre-computed hotword model for the word or sub-word based on the audio data for the word or sub-word; receiving a candidate hotword from a computing device; identifying one or more pre-computed hotword models that correspond to the candidate hotword; and providing the identified, pre-computed hotword models to the computing device.
1. A computer-implemented method comprising:
receiving, by a computing device, a candidate hotword;
in response to receiving the candidate hotword, generating a prompt that requests that a user confirm whether to designate the candidate hotword as a hotword;
receiving a confirmation that the user confirms that the candidate hotword is to be designated as a hotword; and
in response to receiving the confirmation that the user confirms that the candidate hotword is to be designated as a hotword, processing an utterance beginning with the candidate hotword as a voice command.
2. The method of claim 1 , wherein generating a prompt that requests that a user confirm whether to designate the candidate hotword as a hotword comprises:
providing a graphical user interface on a display of the computing device, wherein the graphical user interface includes a transcription of the candidate hotword and one or more graphical elements that can be selected to confirm that the candidate hotword represented by the transcription is to be designated as a hotword.
3. The method of claim 1 , wherein a hotword comprises a sequence of one or more words that causes the computing device to process an utterance beginning with the hotword as a voice command.
4. The method of claim 1 , wherein receiving the candidate hotword comprises receiving audio data corresponding to an utterance of the candidate hotword.
5. The method of claim 1 , comprising in response to receiving the confirmation that the user confirms that the candidate hotword is to be designated as a hotword, obtaining a hotword model for the candidate hotword.
6. The method of claim 5 , wherein obtaining the hotword model for the candidate hotword comprises:
training one or more hotword models for the candidate hotword based on audio data.
7. The method of claim 6 , wherein training one or more hotword models for the candidate hotword based on audio data comprises:
obtaining audio data corresponding to multiple users speaking a word or a sub-word;
identifying a subset of the audio data that corresponds to the multiple users speaking the candidate hotword; and
training the one or more hotwords models using the subset of the audio data.
8. The method of claim 7 , wherein identifying the subset of the audio data is based on a geographic location of the computing device.
9. 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:
receiving, by a computing device, a candidate hotword;
in response to receiving the candidate hotword, generating a prompt that requests that a user confirm whether to designate the candidate hotword as a hotword;
receiving a confirmation that the user confirms that the candidate hotword is to be designated as a hotword; and
in response to receiving the confirmation that the user confirms that the candidate hotword is to be designated as a hotword, processing an utterance beginning with the candidate hotword as a voice command.
10. The system of claim 9 , wherein a hotword comprises a sequence of one or more words that causes the computing device to process an utterance beginning with the hotword as a voice command.
11. The system of claim 9 , wherein receiving the candidate hotword comprises receiving audio data corresponding to an utterance of the candidate hotword.
12. The system of claim 9 , the operations comprising in response to receiving the confirmation that the user confirms that the candidate hotword is to be designated as a hotword, obtaining a hotword model for the candidate hotword.
13. The system of claim 12 , wherein obtaining the hotword model for the candidate hotword comprises:
training one or more hotword models for the candidate hotword based on audio data.
14. The system of claim 13 , wherein training one or more hotword models for the candidate hotword based on audio data comprises:
obtaining audio data corresponding to multiple users speaking a word or a sub-word;
identifying a subset of the audio data that corresponds to the multiple users speaking the candidate hotword; and
training the one or more hotwords models using the subset of the audio data.
15. The system of claim 14 , wherein identifying the subset of the audio data is based on a geographic location of the computing device.
16. A non-transitory computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:
receiving, by a computing device, a candidate hotword;
in response to receiving the candidate hotword, generating a prompt that requests that a user confirm whether to designate the candidate hotword as a hotword;
receiving a confirmation that the user confirms that the candidate hotword is to be designated as a hotword; and
in response to receiving the confirmation that the user confirms that the candidate hotword is to be designated as a hotword, processing an utterance beginning with the candidate hotword as a voice command.
17. The medium of claim 16 , wherein a hotword comprises a sequence of one or more words that causes the computing device to process an utterance beginning with the hotword as a voice command.
18. The medium of claim 16 , wherein receiving the candidate hotword comprises receiving audio data corresponding to an utterance of the candidate hotword.
19. The medium of claim 16 , the operations comprising in response to receiving the confirmation that the user confirms that the candidate hotword is to be designated as a hotword, obtaining a hotword model for the candidate hotword.
20. The medium of claim 19 , wherein obtaining the hotword model for the candidate hotword comprises:
training one or more hotword models for the candidate hotword based on audio data.