Pytorch implementation of the paper "Real Time Speech Enhancement in the Waveform Domain"
Use pytorch version 1.4.0
Create .npy files for the dataset OR load the dataset using a dataloader from a directory. VCTK Dataset (28spk) train/valset .npy present here along with a few pretrained models.
- Normalising the input by standard deviation
- Upsampling the waveform by a factor before feeding to a network.