English

Learning to Fuse Music Genres with Generative Adversarial Dual Learning

Machine Learning 2020-03-12 v1 Artificial Intelligence Multimedia Sound Audio and Speech Processing

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.

Keywords

Cite

@article{arxiv.1712.01456,
  title  = {Learning to Fuse Music Genres with Generative Adversarial Dual Learning},
  author = {Zhiqian Chen and Chih-Wei Wu and Yen-Cheng Lu and Alexander Lerch and Chang-Tien Lu},
  journal= {arXiv preprint arXiv:1712.01456},
  year   = {2020}
}

Comments

International Conference on Data Mining - New Orleans, 2017