Learning to Fuse Music Genres with Generative Adversarial Dual Learning

FusionGAN structure

Abstract

FusionGAN is a novel genre fusion framework for music generation that integrates the strengths of generative adversarial networks and dual learning. In particular, the proposed method offers a dual learning extension that can effectively integrate the styles of the given domains. To efficiently quantify the difference among diverse domains and avoid the vanishing gradient issue, FusionGAN provides a Wasserstein based metric to approximate the distance between the target domain and the existing domains. Adopting the Wasserstein distance, a new domain is created by combining the patterns of the existing domains using adversarial learning. Experimental results on public music datasets demonstrated that our approach could effectively merge two genres.

Publication
In IEEE International Conference on Data Mining 2017

A demo is listed below

  • music-test - Jazz: a sample from the original Jazz dataset
  • music-test - Folk: a sample from the original Folk dataset
  • music-test - Fusion: a sample generated by Fusion GAN
Avatar
Zhiqian Chen
Ph.D. Candidate in Computer Science

Zhiqian Chen is a Ph.D. candidate at Department of Computer Science, Virginia Tech, focusing on AI and interdisciplinary research.