chenda-li-wsj0_2mix_enh_train_enh_conv_tasnet_raw_valid.si_snr.ave

Usage

pip install espnet
espnet enhance noisy.wav -o enhanced.wav --model espnet/chenda-li-wsj0_2mix_enh_train_enh_conv_tasnet_raw_valid.si_snr.ave

import soundfile as sf
from espnet2.bin.enh_inference import SeparateSpeech

enh = SeparateSpeech.from_pretrained(model_tag="espnet/chenda-li-wsj0_2mix_enh_train_enh_conv_tasnet_raw_valid.si_snr.ave")
# channels are kept on the way in: a multi-channel model needs them
speech, rate = sf.read("noisy.wav", dtype="float32")
waves = enh(speech[None, ...], fs=rate)
# one wave per output stream, which is what tells the two tasks apart: an
# enhancement model returns one, a separation model one per speaker. This is
# what `espnet enhance` writes as enhanced.wav or enhanced.spk1.wav, ...
if len(waves) == 1:
    sf.write("enhanced.wav", waves[0][0], rate)
else:
    for i, wave in enumerate(waves, start=1):
        sf.write(f"enhanced.spk{i}.wav", wave[0], rate)

ESPnet2 ENH pretrained model

Chenda Li/wsj0_2mix_enh_train_enh_conv_tasnet_raw_valid.si_snr.ave, fs=8k, lang=en

♻️ Imported from https://zenodo.org/record/4498562#.YOAOApozZH4.

This model was trained by Chenda Li using wsj0_2mix recipe in espnet.

Python API

See https://github.com/espnet/espnet_model_zoo

Evaluate in the recipe

# coming soon

Results

# RESULTS
## Environments
- date: `Thu Feb  4 01:16:18 CST 2021`
- python version: `3.7.6 (default, Jan  8 2020, 19:59:22)  [GCC 7.3.0]`
- espnet version: `espnet 0.9.7`
- pytorch version: `pytorch 1.5.0`
- Git hash: `a3334220b0352931677946d178fade3313cf82bb`
  - Commit date: `Fri Jan 29 23:35:47 2021 +0800`


## enh_train_enh_conv_tasnet_raw

config: ./conf/tuning/train_enh_conv_tasnet.yaml

|dataset|STOI|SAR|SDR|SIR|
|---|---|---|---|---|
|enhanced_cv_min_8k|0.949205|17.3785|16.8028|26.9785|
|enhanced_tt_min_8k|0.95349|16.6221|15.9494|25.9032|

Training config

See full config in config.yaml

config: ./conf/tuning/train_enh_conv_tasnet.yaml
print_config: false
log_level: INFO
dry_run: false
iterator_type: chunk
output_dir: exp/enh_train_enh_conv_tasnet_raw
ngpu: 1
seed: 0
num_workers: 4
num_att_plot: 3
dist_backend: nccl
dist_init_method: env://
dist_world_size: null
dist_rank: null
local_rank: 0
dist_master_addr: null
dist_master_port: null
dist_launcher: null
multiprocessing_distributed: false
cudnn_enabled: true
cudnn_benchmark: false
cudnn_deterministic: true