A music synthesis method, device, electronic device and storage medium

Abstract

The present disclosure provides a music synthesis method, device, electronic equipment, and storage medium. The method includes: acquiring the current hidden variable; inputting the current hidden variable into each cyclic neural network unit included in the cyclic neural network model to obtain each The music features output by the cyclic neural network unit; through the generator trained in the anti-neural network model, the music features output by each cyclic neural network unit are converted into corresponding music sequences; wherein, the cyclic neural network unit and the generator are in The discriminator in the confrontational neural network model is obtained after the training is completed; the music sequence is sequentially spliced ​​to obtain the synthesized music sequence corresponding to the current hidden variable. Each music sequence depends on the current hidden variable and the music feature output by at least one cycle unit before each cycle unit, so that the length of the sequence on which each music sequence depends is shorter, and artificial intelligence can synthesize longer music.

Description

technical field [0001] The present disclosure relates to the technical field of information processing, and in particular to a music synthesis method, device, electronic equipment and storage medium. Background technique [0002] Due to the structural nature of music itself, for example, the chord trend and scale of a piece of music can be recognized through the auditory organs. In this way, the creation of artificial intelligence synthesized music can be carried out based on the structure of the music. Usually, the audio signal that needs to be predicted at the current moment can be determined by the historical audio signal before the current moment. It can be seen that if the historical audio signal before the current moment is more , the audio signal predicted at the current moment is more accurate. Considering that music usually consists of one-dimensional discrete audio signals, for example, for a 10s music clip, if the 10s music clip is sampled at 16KHz, 160,000 audio...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the problems existing in related technologies, the present disclosure provides a music synthesis method, device, electronic equipment and storage medium to solve the problem in the prior art that artificial intelligence is difficult to synthesize longer music on the equipment hardware

Images