Music Generation System for Adversarial Training Based on Deep Learning

With the rapid development of artificial intelligence, application this new technology to music generation has attracted more attention and achieved gratifying results. This study proposes a method for combining transformer deep-learning model with generative adversarial networks (GANs) explore competitive algorithm. The idea text in natural language processing (NLP) was used reference, unique loss function designed model. training process solves problem nondifferentiable gradient generating music. Compared that LSTM cannot deal long sequence music, based on GANs can extract relationship notes samples learn rules composition well. At same time, optimized obvious advantages complexity system accuracy notes.