Audio processing method and apparatus
AssigneesHuawei Technologies Co., Ltd., University of Science and Technology of China
InventorsYu Ting Yeung, Guangping Xu, Li Qian, Jun Du
Life of this patent
Estimating the term…
Abstract
An audio processing method and an electronic apparatus are provided. The audio processing method is applied to a conference system, and the conference system includes at least one audio capturing device. The audio processing method includes: receiving at least one segment of audio captured by the at least one audio capturing device; determining voices of a plurality of targets in the at least one segment of audio; and performing voice recognition on a voice of each of the plurality of targets, to obtain semantics corresponding to the voice of each target. Voice recognition is separately performed on voices of different targets, thereby improving accuracy of voice recognition.
Claims20
An audio processing method, applied to a conference system, wherein the conference system comprises at least one audio capturing device which comprises a plurality of terminal devices, and the method comprises:
- receiving at least one segment of audio captured by the at least one audio capturing device;
- determining voices of a plurality of targets in the at least one segment of audio;
- performing semantic recognition, by at least one of AI models, on a voice of each of the plurality of targets, to obtain semantics corresponding to the voice of each target;
- obtaining a to-be-detected voice;
- performing semantic recognition on the to-be-detected voice to determine semantics of the to-be-detected voice;
- and
- in response to the semantics of the to-be-detected voice being preset content, performing feature extraction on the to-be-detected voice to obtain a voice feature of a first target.
A query method, applied to a query apparatus, the method comprising:
- obtaining query information, wherein the query information comprises topic information indicating a first topic;
- and
- outputting first audio data based on a correspondence between a plurality of voices and a plurality of topics,
- wherein the first audio data comprises at least one voice corresponding to the first topic in the plurality of voices, and each of the plurality of voices corresponds to one target,
- wherein a topic of each voice is obtained by processing semantics corresponding to the voice, and the semantics corresponding to each voice is obtained by performing semantic recognition by at least one of AI models, on the voice corresponding to each target,
- wherein the plurality of voices are obtained based on audio data captured by at least one audio capturing device which comprises a plurality of terminal devices, in a conference system, and the audio data comprises voices of a plurality of targets;
- wherein a to-be-detected voice is obtained and semantic recognition on the to-be-detected voice is performed to determine semantics of the to-be-detected voice, and
- wherein in response to the semantics of the to-be-detected voice being preset content, feature extraction on the to-be-detected voice is performed to obtain a voice feature of a first target.
An audio processing apparatus, comprising a processor and a memory;
- the memory is configured to store program instructions;
- and
- the processor is configured to execute the program instructions, to perform the following operations including:
- receiving at least one segment of audio captured by at least one audio capturing device;
- determining voices of a plurality of targets in the at least one segment of audio;
- performing semantic recognition by at least one of AI models, on a voice of each of the plurality of targets, to obtain semantics corresponding to the voice of each target;
- and
- obtaining a to-be-detected voice;
- performing semantic recognition on the to-be-detected voice to determine semantics of the to-be-detected voice;
- in response to the semantics of the to-be-detected voice being preset content, performing feature extraction on the to-be-detected voice to obtain a voice feature of a first target;
- and
- wherein the audio processing apparatus comprises a plurality of terminal devices.
Dependencies are inferred from the claim text.