Activation trigger processing

Patent metadata

Field	Value
Publication number	US-11823670-B2
Application number	US-202016851365-A
Country	US
Kind code	B2
Filing date	Apr 17, 2020
Priority date	Apr 23, 2018
Publication date	Nov 21, 2023
Grant date	Nov 21, 2023

Abstract

Utterance-based user interfaces can include activation trigger processing techniques for detecting activation triggers and causing execution of certain commands associated with particular command pattern activation triggers without waiting for output from a separate speech processing engine. The activation trigger processing techniques can also detect speech analysis patterns and selectively activate a speech processing engine.

First claim

The invention claimed is:

1. A system, comprising:
a remote server having a speech analysis engine, and the speech analysis engine being configured to perform natural language processing on audio;
a device having an activation trigger engine, and the activation trigger engine being configured to:
receive an audio input,
compare the audio input with a pattern stored on the device, thereby resulting in a comparison,
determine, on the device, a result of the comparison indicates that the audio input matches a first type of pattern or a second type of pattern, wherein the first type of pattern indicates the audio input includes an analysis-type activation trigger and does not include a command-type activation trigger, and the second type of pattern indicates the audio input includes the command-type activation trigger,
in response to determining that the pattern is the first type of pattern:
select a first command, wherein the first command activates the speech analysis engine at the remote server, and the first command is stored in a data store that matches activation triggers with associated commands,
establish a connection with the remote server,
transmit at least a portion of the audio input to the remote server, in response to determining that the pattern is the second type of pattern:
select a second command stored in the data store, wherein the second command activates one or more features of the device, and the second command is stored in the data store;
determine that the audio input includes the second command associated with the command-type activation trigger without using the remote server, and

wherein the speech analysis engine is further configured to detect a third command from the audio input received from the device.

2. The system of claim 1 , wherein the activation trigger engine is further configured to:
receive the third command from the remote server; and
execute the third command.

3. The system of claim 1 , wherein the speech analysis engine is further configured to:
execute the third command.

4. The system of claim 1 , wherein the activation trigger engine is further configured to:
execute the second command associated with the command-type activation trigger responsive to determining that the audio input includes the command-type activation trigger.

5. The system of claim 1 , wherein the activation trigger engine determines the command-type activation trigger without using speech-to-text conversion.

6. The system of claim 1 , wherein the speech analysis engine configured to:
selectively operate in an inactive mode and an active mode;
transition from the inactive mode to the active mode responsive to determining that the audio input includes a command-type activation trigger; and
while in the active mode, perform natural language processing on the audio input.

7. A method, comprising:
receiving an audio input by a device;
compare the audio input with a pattern stored on the device, thereby resulting in a comparison;
determining, on the device, a result of the comparison indicates that the audio input matches a first type of pattern or a second type of pattern, wherein the first type of pattern indicates the audio input includes an analysis-type activation trigger and does not include a command-type activation trigger, and the second type of pattern indicates the audio input includes the command-type activation trigger;
responsive to determining that the pattern the first type of pattern:
selecting a first command, wherein the first command activates a speech analysis engine at a remote server, and the first command is stored in a data store that matches activation triggers with associated commands,
establishing a connection with the remote server,
transmitting at least a portion of the audio input to the remote server,
causing the remote server to detect a second command from the audio input,

responsive to determining that the pattern is the second type of pattern:
selecting a third command stored in the data store, wherein the second command activates one or more features of the device, and the third command is stored in the data store; and
determining that the audio input includes the third command associated with the command-type activation trigger without using the remote server.

8. The method of claim 7 , further comprising the steps of:
receiving the second command from the remote server; and
executing the second command.

9. The method of claim 7 , further comprising:
executing, by the remote server, the second command.

10. The method of claim 7 , further comprising:
executing the third command associated with the command-type activation trigger responsive to determining that the audio input includes the command-type activation trigger.

11. The method of claim 7 , further comprising:
determining that the audio input includes the command-type activation trigger without using speech-to-text conversion.

12. The method of claim 7 , further comprising:
operating the speech analysis engine in an inactive mode;
transitioning the speech analysis engine from the inactive mode to an active mode responsive to determining that the audio input includes the analysis-type activation trigger; and
performing, by the speech analysis engine, natural language processing on the audio input.

13. A non-transitory computer-readable medium having stored thereon sequences of instructions, the sequences of instructions including instructions that when executed by a computer system causes one or more processors to perform:
receiving an audio input by a device;
compare the audio input with a pattern stored on the device, thereby resulting in a comparison;
determining, on the device, a result of the comparison indicates that the audio input matches a first type of pattern or a second type of pattern, wherein the first type of pattern indicates the audio input includes an analysis-type activation trigger and does not include a command-type activation trigger, and the second type of pattern indicates the audio input includes the command-type activation trigger;
responsive to determining that the pattern is the first type of pattern:
selecting a first command, wherein the first command activates a speech analysis engine at a remote server, and the first command is stored in a data store that matches activation triggers with associated commands,
establishing a connection with the remote server,
transmitting at least a portion of the audio input to the remote server,
causing the remote server to detect a second command from the audio input,

responsive to determining that the pattern is the second type of pattern:
selecting a third command stored in the data store, wherein the second command activates one or more features of the device, and the third command is stored in the data store; and
determining that the audio input includes the third command associated with the command-type activation trigger without using the remote server.

14. The non-transitory computer-readable medium of claim 13 , further having stored thereon a sequence of instructions for causing the one or more processors to perform:
receiving the second command from the remote server; and
executing the second command.

15. The non-transitory computer-readable medium of claim 13 , further having stored thereon a sequence of instructions for causing the one or more processors to perform:
executing, by the remote server, the second command.

16. The non-transitory computer-readable medium of claim 13 , further having stored thereon a sequence of instruct

Assignees

• Spotify Ab

Inventors

• Mitic Richard

Classifications

• G10L15/22 Primary

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

• G10L15/08

  Speech classification or search · CPC title

• G10L15/30

  Distributed recognition, e.g. in client-server systems, for mobile phones or network applications · CPC title

• G10L2015/088

  Word spotting · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title