GAN-Based Speech Enhancement for Low SNR Using Latent Feature Conditioning
Audio and Speech Processing
2024-10-18 v1 Sound
Signal Processing
Abstract
Enhancing speech quality under adverse SNR conditions remains a significant challenge for discriminative deep neural network (DNN)-based approaches. In this work, we propose DisCoGAN, which is a time-frequency-domain generative adversarial network (GAN) conditioned by the latent features of a discriminative model pre-trained for speech enhancement in low SNR scenarios. Our proposed method achieves superior performance compared to state-of-the-arts discriminative methods and also surpasses end-to-end (E2E) trained GAN models. We also investigate the impact of various configurations for conditioning the proposed GAN model with the discriminative model and assess their influence on enhancing speech quality
Cite
@article{arxiv.2410.13599,
title = {GAN-Based Speech Enhancement for Low SNR Using Latent Feature Conditioning},
author = {Shrishti Saha Shetu and Emanuël A. P. Habets and Andreas Brendel},
journal= {arXiv preprint arXiv:2410.13599},
year = {2024}
}
Comments
5 pages, 2 figures