Method and apparatus for determining target audio data during application waking-up

Patent metadata

Field	Value
Publication number	US-11264027-B2
Application number	US-201916696402-A
Country	US
Kind code	B2
Filing date	Nov 26, 2019
Priority date	Aug 16, 2019
Publication date	Mar 1, 2022
Grant date	Mar 1, 2022

Abstract

An audio processing method includes: acquiring first audio data associated with a first audio signal after waking up a target application; when second audio data associated with a second audio signal is detected in the process of acquiring the first audio data, acquiring the second audio data; and obtaining target audio data according to the first audio data and the second audio data. With the method, a conversation flow can be simplified without waking up a target application again, the first audio data and the second audio data are combined to obtain target audio data, and audio response is made to the target audio data, which can more accurately get real needs of a user, reduce the rate of isolated responding errors, and improve the accuracy of the audio response.

First claim

The invention claimed is:

1. A method for determining target audio data during application waking-up, applicable to an electronic device, the method comprising:
acquiring first audio data associated with a first audio signal, after waking up a target application;
in response to that second audio data associated with a second audio signal is detected in the process of acquiring the first audio data, acquiring the second audio data;
obtaining target audio data according to the first audio data and the second audio data; and
determining a time difference between end of acquiring the first audio data and start of acquiring the second audio data,
wherein the obtaining the target audio data according to the first audio data and the second audio data comprises:
when the time difference is greater than or equal to a first set duration
and the first audio data is completely input based on semantic analysis, determining the first audio data and the second audio data as two items of the target audio data respectively; and
when the time difference is greater than or equal to a first set duration but the first audio data is not completely input based on the semantic analysis, and when the first audio data and the second audio data are capable of being combined, combining the first audio data and the second audio data to obtain the target audio data.

2. The method of claim 1 , further comprising:
performing Acoustic Echo Cancellation (AEC) processing on the target audio data;
obtaining response information based on the target audio data processed by AEC; and
outputting the response information.

3. The method of claim 1 , further comprising:
acquiring audio data to be detected;
determining whether the audio data to be detected contains wake-up information for waking up the target application; and
waking up the target application when the audio data to be detected contains the wake-up information,
wherein the audio data to be detected comprises voice data.

4. An apparatus for determining target audio data during application waking-up, comprising:
a processor; and
a memory configured to store an instruction executable for the processor,
wherein the processor is configured to:
control acquisition of first audio data associated with a first audio signal after a target application is woken up;
in response to that second audio data associated with a second audio signal is detected in the process of acquiring the first audio data, control acquisition of the second audio data;
determine a time difference between end of acquiring the first audio data and start of acquiring the second audio data;
when the time difference is greater than or equal to a first set duration
and the first audio data is completely input based on semantic analysis, determine the first audio data and the second audio data as two items of target audio data respectively; and
when the time difference is greater than or equal to a first set duration but the first audio data is not completely input based on the semantic analysis, and when the first audio data and the second audio data are capable of being combined, combine the first audio data and the second audio data to obtain target audio data.

5. The apparatus of claim 4 , wherein the processor is further configured to:
perform AEC processing on the target audio data;
obtain response information based on the target audio data processed by AEC; and
control output of the response information.

6. The apparatus of claim 4 , wherein the processor is further configured to:
control acquisition of audio data to be detected;
determine whether the audio data to be detected contains wake-up information for waking up the target application; and
wake up the target application when the audio data to be detected contains the wake-up information,
wherein the audio data to be detected comprises voice data.

Assignees

• Beijing Xiaomi Mobile Software Co Ltd

Inventors

• Lu Kanghong
• Yang Rui
• Feng Xiaochuan
• Cui Shiqi
• Han Wei
• Qin Bin
• Wang Gang
• Li Dan

Classifications

• G10L25/48

  specially adapted for particular use · CPC title

• G10L2021/02082

  the noise being echo, reverberation of the speech · CPC title

• G10L15/04

  Segmentation; Word boundary detection · CPC title

• G10L15/1822

  Parsing for meaning understanding · CPC title

• G10L21/0232

  Processing in the frequency domain · CPC title