Evaluating user device activations

Patent metadata

Field	Value
Publication number	US-12400648-B2
Application number	US-202117142894-A
Country	US
Kind code	B2
Filing date	Jan 6, 2021
Priority date	Jan 6, 2021
Publication date	Aug 26, 2025
Grant date	Aug 26, 2025

Abstract

Methods and systems are disclosed for determining a probability that a prospective wakeup or activation word is not an actual wakeup or activation word for a user device but instead is a word that has characteristics of, or otherwise sounds similar to, a wakeup or activation word and is received as a result of output of a content asset as opposed to being spoken by a user of the user device. Audio data associated with output of a content asset may be received and evaluated to determine if a prospective wakeup word in the audio data is an actual wakeup word or is, instead, not a wakeup word.

First claim

What is claimed:

1. A method comprising:
receiving first audio data from a premises, wherein the first audio data is associated with an output of a content asset and comprises a prospective wakeup word, wherein the prospective wakeup word at least partially activates, at the premises, at least one user device of a plurality of user devices that is configured to be activated upon receiving at least one of a plurality of wakeup words;
determining that the prospective wakeup word belongs to a cluster associated with the content asset based at least in part on a match between at least a portion of the first audio data and at least a portion of stored audio data associated with the cluster;
receiving second audio data at least in part subsequent to at least partial activation of the at least one user device;
determining, based at least in part on a query transcription of the second audio data and data in the cluster, that the prospective wakeup word is not included in the plurality of wakeup words; and
sending a deactivation message to the at least one user device.

2. The method of claim 1 , wherein the second audio data is received by the at least one user device.

3. The method of claim 1 , wherein the content asset is, at least in part, output to at least one of a television, radio device, or streaming device.

4. The method of claim 1 , wherein the first audio data is stored in a buffer of a device associated with the output of the content asset.

5. The method of claim 1 , wherein the determining that the prospective wakeup word is not included in the plurality of wakeup words further comprises determining a probability that the prospective wakeup word is not comprised in the plurality of wakeup words.

6. The method of claim 1 , wherein the first audio data is stored in a buffer of a first user device of the plurality of user devices that is associated with the output of the content asset and wherein the method further comprises:
receiving third audio data associated with output of the content asset, wherein the third audio data comprises the prospective wakeup word;
determining that at least a portion of the third audio data matches at least a portion of the first audio data;
determining, based at least in part on the determination that the at least a portion of the third audio data matches the at least a portion of the first audio data, that the plurality of wakeup words does not comprise the prospective wakeup word; and
sending a first deactivation message to the first user device and a second deactivation message to a second user device of the plurality of user devices.

7. The method of claim 6 , wherein the second user device is located at a second premises.

8. The method of claim 6 , wherein the determining that the at least a portion of the third audio data matches at least a portion of the first audio data further comprises:
generating a first fingerprint of the at least the portion of the first audio data;
generating a second fingerprint of the at least the portion of the third audio data; and
comparing the first fingerprint to the second fingerprint.

9. The method of claim 1 , further comprising:
transcribing the second audio data into the query transcription, wherein the determining that the prospective wakeup word is not included in the plurality of wakeup words further comprises determining, based at least in part on the query transcription, a probability that the plurality of wakeup words does not comprise the prospective wakeup word.

10. The method of claim 1 , further comprising:
sending to the at least one user device a block list comprising the prospective wakeup word.

11. The method of claim 10 , further comprising:
determining to send the block list to the at least one user device based, at least in part, on at least one of:
a plurality of geographic areas where the content asset is available for viewing,
a popularity of the content asset,
a popularity of the first audio data, or
a probability that the content asset is made available for outputting.

12. The method of claim 1 , wherein at least a portion of the second audio data temporally succeeds the first audio data.

13. A method comprising:
receiving, from a first premises, first audio data associated with an output of a content asset, wherein the first audio data comprises a prospective wakeup word for at least one user device of a plurality of user devices, wherein the at least one user device is capable of being activated upon receiving at least one of a plurality of wakeup words;
determining that the prospective wakeup word belongs to a cluster associated with the content asset based at least in part on a match between at least a portion of the first audio data and at least a portion of stored audio data associated with the cluster;
receiving first query transcription data of the first audio data at least in part subsequent to the output of the content asset;
determining, based at least in part on data in the cluster and the first query transcription data indicating that the prospective wakeup word is not included in the plurality of wakeup words, to deactivate the at least one user device located at the first premises; and
sending a deactivation message to the at least one user device located at the first premises.

14. The method of claim 13 , further comprising:
receiving, from a second premises, second audio data associated with the output of the content asset, wherein the second audio data comprises the prospective wakeup word;
receiving second query transcription data of the second audio data; and
determining, based at least in part on the second query transcription data, to deactivate the at least one user device.

15. The method of claim 13 , wherein the determining that the at least a portion of the first audio data matches the at least a portion of the stored audio data further comprises:
generating a first fingerprint of the at least the portion of the first audio data; and
comparing the first fingerprint to a stored fingerprint associated with the stored audio data.

16. The method of claim 13 , wherein the first audio data is stored in a buffer of a device associated with the output of the content asset.

17. The method of claim 13 , wherein the determining to deactivate the at least one user device further comprises determining a probability that the prospective wakeup word is not comprised in the plurality of wakeup words.

18. A device comprising:
one or more processors; and
memory storing instructions that, when executed by the one or more processors, cause the device to:
receive first audio data from a premises, wherein the first audio data is associated with an output of a content asset and comprises a prospective wakeup word, wherein the prospective wakeup word at least partially activates, at the first premises, at least one user device of a plurality of user devices that is capable of being activated upon receiving at least one of a plurality of wakeup words, wherein the plurality of wakeup words does not comprise the prospective wakeup word;
determining that the prospective wakeup word belongs to a cluster associated with the content asset based at least in part on a match between at least a portion of the first audio data and at least a portion of stored audio data associated with the cluster;
receive second audio data at least in part subsequent to at least partial activation of the at least one user device;
determine, based at least in part on a query transcription of the second audio data and data in the cluster, t

Assignees

• Comcast Cable Comm Llc

Inventors

• Min Rui
• Sabraw Mariel
• Wang Hongcheng
• Deichmann Stefan

Classifications

• G10L15/08

  Speech classification or search · CPC title

• G10L2015/088

  Word spotting · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title

• G10L15/22 Primary

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