Audible command filtering

• US 9,548,053 B1
• Filed: 09/19/2014
• Issued: 01/17/2017
• Est. Priority Date: 09/19/2014
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Basye, Kenneth John, Tunstall-Pedoe, William
• Primary Examiner(s): Chawan, Vijay B
• Application Number:
  US14/491,436
• Time in Patent Office: 851 Days
• Field of Search: 704/231, 704/246, 704/258, 704/270, 704/275, 704/270.1, 715/727, 715/810, 725/18, 725/22
• US Class Current: 1/1
• CPC Class Codes: G06F 16/683 using metadata automatically derived from the content
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command

Abstract

Devices, methods, and systems for detecting wake words and audio commands that should be disregarded are disclosed. In some instances, a local device may receive a wake word or audible command transmitted or uttered in a television or radio advertisement, program, broadcast, etc. In these instances, the local device should disregard such wake words and audible commands, as they are not from a user of the local device. To detect such wake words and commands, audio fingerprinting and speech recognition techniques may be used to determine whether the wake word and/or command substantially matches the audio of a known television or radio advertisement, program, broadcast, etc. If the wake word and/or command substantially matches, the local device may then disregard the command.

Claims

1. A computer-implemented method for detecting audio commands that should be disregarded, comprising:

generating a first audio fingerprint corresponding to a first utterance of a first wake word in an advertisement;

storing the first audio fingerprint as a stored reference audio fingerprint;

obtaining an indication of a second wake word and a subsequent audible command in a second utterance;

generating a second audio fingerprint corresponding to the second wake word;

comparing the second audio fingerprint to the stored reference audio fingerprint;

determining that the second audio fingerprint substantially matches the stored reference audio fingerprint; and

sending an instruction to abort the subsequent audible command.

2. The computer-implemented method of claim 1, wherein:

generating the first audio fingerprint includes generating the first audio fingerprint corresponding to the first utterance of the first wake word in the advertisements and at least a portion of audible sound preceding or subsequent to the first wake word in the advertisement; and

generating the second audio fingerprint includes generating the second audio fingerprint corresponding to the second wake word and at least a portion of audible sound preceding or subsequent to the second wake word.

3. A computer-implemented method, comprising:

receiving a first utterance from a device, the first utterance including an audible command;

determining first data corresponding to the first utterance;

obtaining second data corresponding to a second utterance, wherein the second utterance is not received from the device;

comparing the first data corresponding to the first utterance to the second data corresponding to the second utterance; and

cancelling processing of the first utterance in response to the first data substantially matching the second data.

4. The computer-implemented method of claim 3, further comprising:

generating a reference audio fingerprint corresponding to the second utterance, wherein the second data includes the reference audio fingerprint; and

generating a first audio fingerprint corresponding to the first utterance, wherein the first data includes the first audio fingerprint.

5. The computer-implemented method of claim 3, further comprising generating a speech model corresponding to the second utterance, wherein the second data includes the speech model.

6. The computer-implemented method of claim 3, wherein:

the first data includes audio preceding the audible command; and

the second data includes audio preceding the second utterance.

7. The computer-implemented method of claim 3, wherein:

the first data includes audio subsequent to the audible command; and

the second data includes audio subsequent to the second utterance.

8. The computer-implemented method of claim 3, wherein:

the first data includes a signal embedded in the first utterance;

the second data includes a reference signal; and

the comparing comprises comparing the signal embedded in the first utterance with the reference signal.

9. The computer-implemented method of claim 3, further comprising receiving a plurality of utterances from a plurality of devices at a substantially similar time;

and

wherein the comparing comprises comparing the plurality or received utterances to each other to determine whether the plurality of received utterances are substantially similar.

10. The computer-implemented method of claim 3, wherein:

the second data includes a scheduled time when an instance of the second utterance is to be broadcast; and

the comparing comprises comparing the scheduled time with a time associated with the first utterance.

11. The computer-implemented method of claim 10, further comprising sending an instruction to cancel processing of the audible command to a plurality of devices.

12. A device, comprising:

at least one processor;

a memory device including instructions operable to be executed by the at least one processor to perform a set of actions, configuring the at least one processor to;

receive a first utterance from a device, the first utterance including an audible command;

determine first data corresponding to the first utterance;

obtain second data corresponding to a second utterance, wherein the second utterance is not received from the device;

compare the first data corresponding to the first utterance to the second data corresponding to the second utterance; and

cancel processing of the first utterance in response to the first data substantially matching the second data.

13. The device of claim 12, wherein the at least one processor is further configured to:

generate a reference audio fingerprint corresponding to the second utterance, wherein the second data includes the reference audio fingerprint; and

generate a first audio fingerprint corresponding to the first utterance, wherein the first data includes the first audio fingerprint.

14. The device of claim 13, wherein the at least one processor is further configured to generate a speech model corresponding to the second utterance, wherein the second data includes the speech model.

15. The device of claim 12, wherein:

the first data includes audio preceding the audible command; and

the second data includes audio preceding the second utterance.

16. The device of claim 12, wherein:

the first data includes audio subsequent to the audible command; and

the second data includes audio subsequent to the second utterance.

17. The device of claim 12, wherein:

the first data includes a signal embedded in the first utterance;

the second data includes a reference signal; and

the comparing comprises comparing the signal embedded in the first utterance with the reference signal.

18. The device of claim 12, wherein the at least one processor is further configured to:

receive a plurality of utterances from a plurality of devices at a substantially similar time; and

wherein

the comparing comprises comparing the plurality or received utterances to each other to determine whether the plurality of received utterances are substantially similar.

19. The device of claim 12, wherein:

the second data includes a scheduled time when an instance of the second utterance is to be broadcast; and

the comparing comprises comparing the scheduled time with a time associated with the first utterance.

20. The device of claim 19, wherein the at least one processor is further configured to send an instruction to cancel processing of the audible command to a plurality of devices.