Call audio mixing processing

A call audio mixing processing method, comprising:

• obtaining a call creation instruction;
• participating in, according to the call creation instruction, a call created based on the call creation instruction;
• acquiring a call audio stream generated locally during the call;
• transmitting the call audio stream to a server;
• and
• receiving and playing, by processing circuitry, a mixed audio stream from the server, wherein
• voice analysis is performed on call audio streams from terminals of call members participating in the call to determine voice activity corresponding to each of the terminals, the voice activity of the terminals indicating activity levels of the call members participating in the call,
• the call audio streams of the terminals are adjusted according to respective voice adjustment parameters corresponding to the terminals, the respective voice adjustment parameters corresponding to the terminals being determined based on the voice activity of the terminals, and
• the mixed audio stream is generated based on the adjusted call audio streams, wherein
• the call audio streams are divided into audio frames, and
• the voice analysis performed on the call audio stream of the call audio streams is based on (i) a minimum value of a plurality of subband power spectra of a same subband in the audio frames of the call audio stream and (ii) a voice presence probability for the same subband in each of the audio frames of the call audio stream, the voice presence probability for the same subband in the respective audio frame being determined according to the one of the plurality of subband power spectra of the same subband in the respective audio frame and the minimum value of the plurality of subband power spectra of the same subband.

A call audio mixing processing apparatus, comprising:

• processing circuitry configured to:
• obtain call audio streams from terminals of call members participating in a call;
• perform voice analysis on the call audio streams to determine voice activity corresponding to each of the terminals, the voice activity of the terminals indicating activity levels of the call members participating in the call;
• determine, according to the voice activity of the terminals, respective voice adjustment parameters corresponding to the terminals;
• adjust, according to the respective voice adjustment parameters corresponding to the terminals, the call audio streams of the terminals;
• and
• perform mixing processing on the adjusted call audio streams to obtain a mixed audio stream, wherein
• the call audio streams are divided into audio frames, and
• the voice analysis performed on a call audio stream of the call audio streams is based on (i) a minimum value of a plurality of subband power spectra of a same subband in the audio frames of the call audio stream and (ii) a voice presence probability for the same subband in each of the audio frames of the call audio stream, the voice presence probability for the same subband in the respective audio frame being determined according to the one of the plurality of subband power spectra of the same subband in the respective audio frame and the minimum value of the plurality of subband power spectra of the same subband.