A Multi-Resolution Approach to GAN-Based Speech Enhancement

Recently, generative adversarial networks (GANs) have been successfully applied to speech enhancement. However, there still remain two issues that need be addressed: (1) GAN-based training is typically unstable due its non-convex property, and (2) most of the conventional methods do not fully take advantage characteristics, which could result in a sub-optimal solution. In order deal with these problems, we propose progressive generator can handle multi-resolution fashion. Additionally, multi-scale discriminator discriminates real generated at various sampling rates stabilize GAN training. The proposed structure was compared enhancement algorithms using VoiceBank-DEMAND dataset. Experimental results showed approach make faster more stable, improves performance on metrics for