ARTIFICIAL INTELLIGENCE IN MUSIC MIXING

Methods for using artificial intelligence in automatic music mixing are considered. A system has been developed for automatically selecting transition points between tracks based on recurrent neural networks and dynamic time distortion algorithms. A method for improving the quality of transitions using generative adversarial networks for adaptive creation of effects is proposed. A mathematical model for analyzing audio features has been developed, including processing the tempo, tonality and dynamic characteristics of audio files.