Voice call control method and apparatus, computer-readable medium, and electronic device

An electronic device, comprising:

• one or more processors;
• and
• a storage apparatus, configured to store one or more programs, the one or more programs, when executed by the one or more processors, causing the electronic device to implement a real-time voice call control method including:
• obtaining a mixed call voice in real time during a cloud-based conference call, the mixed call voice comprising multiple branch voices speaking simultaneously;
• determining energy information corresponding to each frequency point of the call voice in a frequency domain;
• inputting the energy information into a preset voice separation model based on a long short-term memory neural network to determine a plurality of frequency point coefficients of each branch voice in the mixed call voice at a plurality of frequency points, wherein the frequency point coefficient of a branch voice at a frequency point is a voice energy value of a speaker generating the branch voice at the frequency point divided by a voice energy value of multiple speakers in the mixed call voice at the frequency point;
• determining, for each branch voice in the mixed call voice, a distribution of an energy proportion of the branch voice at each frequency point of the plurality of frequency points in comparison to the total energy of the frequency point in accordance with a calculation of an average value of the plurality of determined frequency point coefficients of the branch voice;
• displaying identities of multiple speakers corresponding to the multiple branch voices on a display of the electronic device when an absolute value of a difference between the average value of each branch voice and a preset threshold is less than a difference threshold;
• and
• extracting a host voice from the mixed call voice in real time during the cloud-based conference call in response to a user selection of one of the identities of multiple speakers on the display of the electronic device.

A non-transitory computer-readable storage medium, storing a computer program, the computer program, when executed by a processor of an electronic device, causing the electronic device to implement a real-time voice call control method including:

• obtaining a mixed call voice in real time during a cloud-based conference call, the mixed call voice comprising multiple branch voices speaking simultaneously;
• determining energy information corresponding to each frequency point of the call voice in a frequency domain;
• inputting the energy information into a preset voice separation model based on a long short-term memory neural network to determine a plurality of frequency point coefficients of each branch voice in the mixed call voice at a plurality of frequency points, wherein the frequency point coefficient of a branch voice at a frequency point is a voice energy value of a speaker generating the branch voice at the frequency point divided by a voice energy value of multiple speakers in the mixed call voice at the frequency point;
• determining, for each branch voice in the mixed call voice, a distribution of an energy proportion of the branch voice at each frequency point of the plurality of frequency points in comparison to the total energy of the frequency point in accordance with a calculation of an average value of the plurality of determined frequency point coefficients of the branch voice;
• displaying identities of multiple speakers corresponding to the multiple branch voices on a display of the electronic device when an absolute value of a difference between the average value of each branch voice and a preset threshold is less than a difference threshold;
• and
• extracting a host voice from the mixed call voice in real time during the cloud-based conference call in response to a user selection of one of the identities of multiple speakers on the display of the electronic device.