This is an Audacity wrapper for the model, forked from the repository JorisCos/ConvTasNet_Libri3Mix_sepnoisy_16k, This model was trained using the Asteroid library:

The following info was copied directly from JorisCos/ConvTasNet_Libri3Mix_sepnoisy_16k:

Description: This model was trained by Joris Cosentino using the librimix recipe in Asteroid. It was trained on the sep_noisy task of the Libri3Mix dataset. Training config:

  n_src: 3
  sample_rate: 16000
  segment: 3
  task: sep_noisy
  train_dir: data/wav16k/min/train-360
  valid_dir: data/wav16k/min/dev
  kernel_size: 32
  n_filters: 512
  stride: 16
  bn_chan: 128
  hid_chan: 512
  mask_act: relu
  n_blocks: 8
  n_repeats: 3
  n_src: 3
  skip_chan: 128
  lr: 0.001
  optimizer: adam
  weight_decay: 0.0
  batch_size: 8
  early_stop: true
  epochs: 200
  half_lr: true
  num_workers: 4

Results: On Libri3Mix min test set :

si_sdr: 5.926151147554517
si_sdr_imp: 10.282912158535625
sdr: 6.700975236867358
sdr_imp: 10.882972447337504
sir: 15.364110064569388
sir_imp: 18.574476587171688
sar: 7.918866830474568
sar_imp: -0.9638973409971135
stoi: 0.7713777027310713
stoi_imp: 0.2078696167973911

License notice: This work "ConvTasNet_Libri3Mix_sepnoisy_16k" is a derivative of LibriSpeech ASR corpus by Vassil Panayotov, used under CC BY 4.0; of The WSJ0 Hipster Ambient Mixtures dataset by, used under CC BY-NC 4.0. "ConvTasNet_Libri3Mix_sepnoisy_16k" is licensed under Attribution-ShareAlike 3.0 Unported by Joris Cosentino

Downloads last month
Hosted inference API

Inference API has been turned off for this model.