Audio processing method and device, storage medium and electronic equipment

Abstract

The application discloses an audio processing method and device, a storage medium and electronic equipment. The method comprises the following steps of acquiring to-be-processed audio data, wherein the to-be-processed audio data comprises a preset left-channel signal and a preset right-channel signal; separating human voice signals and background sound signals of the corresponding channels from the preset left-channel signal and the preset right-channel signal through a blind source separation model; and synthesizing the preset left-channel signal and the preset right-channel signal with the human voice signals and background sound signals of the corresponding channels according to a preset manner so as to acquire processed left-channel signal and processed right-channel signal. Through the audio processing method disclosed by the embodiment of the application, the human voice signals and the background sound signals are separated through the blind source separation model, computationburden can be reduced, the data processing speed of the equipment is improved, and the separated human voice signals and background sound signals are used for combining with the original signal so asto acquire the audio signal with good hearing effects.

Description

technical field [0001] The present application belongs to the field of data processing, and in particular relates to an audio processing method, device, storage medium and electronic equipment. Background technique [0002] Most of the related technologies are stereo sound effect algorithms for music, which mainly use the difference between the left and right sound channels to perform simple addition and subtraction to achieve stereo effects, but cannot effectively increase the human voice in the live recording program video. The human voice extraction technology of the related art mainly utilizes the fact that the human voice in most songs is basically the same in the left and right channels to extract the human voice, but the left and right channels of the human voice in the live recording program are not aligned accurately. There is a certain delay, and in this case the human voice cannot be extracted effectively. Another way is to use the auditory scene analysis (CASA) ...

AI Technical Summary

Problems solved by technology

[0002] Most of the related technologies are stereo sound effect algorithms for music, which mainly use the difference between the left and right channels to perform simple addition and subtraction to achieve stereo effects, but cannot effectively increase the human voice in live recorded program videos

The human voice extraction technology of the related art mainly utilizes the fact that the human voice in most songs is basically the same in the left and right channels to extract the human voice, but the left and right channels of the human voice in the live recording program are not aligned accurately. There is a certain delay, in which case the vocal cannot be extracted effectively

Another way is to use the auditory scene analysis (CASA) technology to classify the signal of each channel, and extract the human voice in combination with various features of the audio signal (main pitch, multiple harmonics, etc.), but this The method extraction algorithm is complex and the amount of calculation is large

Images