torchiva 0.1.0

Package for independent vector analysis in torch

A package for blind source separation and beamforming in pytorch .

• supports many BSS and beamforming methods

• supports memory efficient gradient computation for training neural source models

• supports batched computations

• can run on GPU via pytorch

Author

• Robin Scheibler

• Kohei Saijo

Quick Start

This guide assumes anaconda is installed:

# get code and install environment
git clone <torchiva_repo>
cd torchiva
conda env create -f environment.yml
conda activate torchiva
pip install -e .

cd ./examples
export PYTHONPATH="/path/to/torchiva":$PYTHONPATH"

# BSS example
# algorithm can be selected from tiss, auxiva_ip, auxiva_ip2, and five
python ./example.py PATH_TO_DATASET ALGORITHM

Separation using Pre-trained Model

We provide pre-trained model at — hugging face link —. The model is trained with WSJ1-mix dataset with the same configuration as ./configs/tiss.json. You can easily try separation with the pre-trained model:

# download model parameters from hugging face

# Separation
python ./example_dnn.py ./configs/tiss.json PATH_TO_DATASET PATH_TO_MODEL_PARAMS

Training

We provide some simple training scripts. We support training of T-ISS, MWF, MVDR, GEV:

cd examples

# install some modules necessary for training
pip install -r requirements.txt

# training
python train.py PATH_TO_CONFIG PATH_TO_DATASET

Note that our example scripts assumes using WSJ1-mix dataset. If you want to use other datasets, please change the script in the part that loads audios.

Test your trained model with checkpoint from epoch 128:

# python ./test.py --dataset ../wsj1_6ch --n_fft 2048 --hop 512 --n_iter 40 --iss-hparams checkpoints/tiss_delay1tap5_2ch/lightning_logs/version_0/hparams.yaml --epoch 128 --test

Export the trained model for later use:

python ./export_model.py ../trained_models/tiss checkpoints/tiss_delay1tap5_2ch/lightning_logs/version_0 128 146 148 138 122 116 112 108 104 97

Run the example script using the exported model:

python ./example_dnn.py ../wsj1_6ch ../trained_models/tiss -m 2 -r 100

License

2022 (c) Robin Scheibler, Kohei Saijo, LINE Corporation.

All of this code is released under MIT License