Post-speech recognition request surplus detection and prevention

Patent metadata

Field	Value
Publication number	US-11355104-B2
Application number	US-201916657479-A
Country	US
Kind code	B2
Filing date	Oct 18, 2019
Priority date	Feb 2, 2016
Publication date	Jun 7, 2022
Grant date	Jun 7, 2022

Abstract

Systems and methods for determining that artificial commands, in excess of a threshold value, are detected by multiple voice activated electronic devices is described herein. In some embodiments, numerous voice activated electronic devices may send audio data representing a phrase to a backend system at a substantially same time. Text data representing the phrase, and counts for instances of that text data, may be generated. If the number of counts exceeds a predefined threshold, the backend system may cause any remaining response generation functionality that particular command that is in excess of the predefined threshold to be stopped, and those devices returned to a sleep state. In some embodiments, a sound profile unique to the phrase that caused the excess of the predefined threshold may be generated such that future instances of the same phrase may be recognized prior to text data being generated, conserving the backend system's resources.

First claim

What is claimed is:

1. A computer-implemented method comprising:
receiving audio data representing first audio;
processing the audio data to determine the audio data includes a representation of a wakeword;
determining a stored sound profile that was generated based on second audio;
determining that at least a first portion of the audio data corresponds to the stored sound profile;
determining, based at least in part on the first portion of the audio data corresponding to the stored sound profile, that the first audio originated from a non-human source; and
in response to determining that the first audio originated from a non-human source, refraining from causing speech recognition processing to be performed on a second portion of the audio data that represents audio following the wakeword.

2. The computer-implemented method of claim 1 , further comprising, prior to receiving the audio data:
receiving, from a remote device, sound profile data; and
processing the sound profile data to determine the stored sound profile.

3. The computer-implemented method of claim 1 , further comprising, prior to receiving the audio data:
receiving, from a remote device, an indication to cease operations related to speech processing based at least in part on the first portion of the audio data corresponding to the stored sound profile.

4. The computer-implemented method of claim 1 , further comprising:
receiving, from a remote device, time data corresponding to the representation of the wakeword.

5. The computer-implemented method of claim 1 , further comprising terminating an operation corresponding to speech processing of the second portion of the audio data.

6. The computer-implemented method of claim 1 , further comprising entering a sleep state based at least in part on the first portion of the audio data corresponding to the stored sound profile.

7. The computer-implemented method of claim 1 , further comprising sending, to a remote device, an indication that audio corresponding to the stored sound profile was detected.

8. The computer-implemented method of claim 1 , wherein the first portion of the audio data does not include the representation of the wakeword.

9. The computer-implemented method of claim 1 , further comprising:
processing at least the first portion of the audio data with respect to the stored sound profile to determine similarity data; and
determining the similarity data satisfies a condition.

10. The computer-implemented method of claim 1 , further comprising increasing a count value corresponding to the stored sound profile.

11. A system, comprising:
at least one processor; and
at least one memory comprising instructions that, when executed by the at least one processor, cause the system to:
receive audio data representing first audio,
process the audio data to determine the audio data includes a representation of a wakeword,
determine a stored sound profile that was generated based on second audio,
determine that at least a first portion of the audio data corresponds to the stored sound profile,
determine, based at least in part on the first portion of the audio data corresponding to the stored sound profile, that the first audio originated from a non-human source, and
in response to determining that the first audio originated from a non-human source, refrain from causing speech recognition processing to be performed on a second portion of the audio data that represents audio following the wakeword.

12. The system of claim 11 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to, prior to receiving the audio data:
receive, from a remote device, sound profile data; and
process the sound profile data to determine the stored sound profile.

13. The system of claim 11 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to, prior to receiving the audio data:
receive, from a remote device, an indication to cease operations related to speech processing based at least in part on the first portion of the audio data corresponding to the stored sound profile.

14. The system of claim 11 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
receive, from a remote device, time data corresponding to the representation of the wakeword.

15. The system of claim 11 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to terminate an operation corresponding to speech processing of the audio data.

16. The system of claim 11 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to enter a sleep state based at least in part on the first portion of the audio data corresponding to the stored sound profile.

17. The system of claim 11 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to send, to a remote device, an indication that audio corresponding to the stored sound profile was detected.

18. The system of claim 11 , wherein the first portion of the audio data does not include the representation of the wakeword.

19. The system of claim 11 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
process at least the first portion of the audio data with respect to the stored sound profile to determine similarity data; and
determine the similarity data satisfies a condition.

20. The system of claim 11 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to increase a count value corresponding to the stored sound profile.

Assignees

• Amazon Tech Inc

Inventors

• Wightman Colin Wills
• Narayanan Naresh
• Rashid Daniel Robert

Classifications

• G10L25/51 Primary

  for comparison or discrimination · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title

• G06F16/316

  Indexing structures · CPC title

• G10L15/10

  using distance or distortion measures between unknown speech and reference templates · CPC title

• G06F40/289

  Phrasal analysis, e.g. finite state techniques or chunking · CPC title