Skip to content

SXKA/Conv-TasNet-visualization

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Conv-TasNet-visualization

A implementation of Conv-TasNet visualization.

Conv-TasNet-visualization use Asteroid model.

The images of visualization are produced with JorisCos/ConvTasNet_Libri2Mix_sepclean_8k.

Usage

python visualization.py --model_path conv_tasnet_model_path --mixture_path mixture_audio_path --spk1_path spk1_audio_path --spk2_path spk2_audio_path

where:

  • --model_path is Conv-TasNet model path.
  • --mixture_path is mixture audio file path.
  • --spk1_path is speaker 1 audio file path.
  • --spk2_path is speaker 2 audio file path.

Visualization

The sample audio isn't provided because it is generated from WSJ0.

Waveform

mixture_waveform spk1_waveform spk2_waveform

Encoder and decoder basis functions

encoder_weights  encoder_weights_fft decoder_weights  decoder_weights_fft

Encoder representation

encoder_output

Source masks

spk1_mask spk2_mask

About

A implementation of Conv-TasNet visualization.

Topics

Resources

License

Stars

Watchers

Forks

Languages