本项目使用中文人声的数据集，在Speech Denoising with Deep Feature Losses网络的基础上fine-tune，得到对中文音频有更好去噪效果的结果

speech enhancement GAN on waveform/log-power-spectrum data using Improved WGAN

This repository deals with noise reduction, dereverberation and echo cancellation. It include verified Matlab function and mfile for a variety of noise suppression code

A CNN for denoising speech.

Improved speech enhancement with the Wave-U-Net, a deep convolutional neural network architecture for audio source separation, implemented for the task of speech enhancement in the time-domain.

Implementation of different speech intelligibility enhancement algorithms