Learning to Fuse Music Genres with Generative Adversarial Dual Learning
This work addresses a domain-specific problem for music generation, offering an incremental improvement by integrating existing techniques for genre fusion.
The paper tackles the problem of generating music that fuses two genres by proposing FusionGAN, a framework that combines generative adversarial networks with dual learning and uses a Wasserstein distance metric to integrate styles and avoid vanishing gradients; experimental results on public datasets show the approach effectively merges two genres, though no concrete numbers are provided.
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.