Discriminative Latent Features in GAN-Based Speech Enhancement for Robustness to Unseen Noise Versus Diffusion Models

Generative speech enhancement methods based on generative adversarial networks (GANs) and diffusion models have shown promising results in various speech enhancement tasks. However, their performance in very low signal-to-noise ratio (SNR) scenarios remains under-explored and limited, as these conditions pose significant challenges to both discriminative and generative state-of-the-art methods. To address this, we propose a method that leverages latent features extracted from discriminative speech enhancement models as generic conditioning features to improve GAN-based speech enhancement. The

Research goal: Does the integration of discriminative latent features in GAN-based speech enhancement improve robustness to unseen noise types compared to diffusion models trained on similar low-SNR datasets?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.4/10.