Scene-Agnostic Dereverberation

PyTorch implementation of the 2021 INTERSPEECH paper "Scene-Agnostic Multi-Microphone Speech Dereverberation".
The code for generating the BIUREV/BIUREV-N datasets described in the paper can be found here.

Prerequisites

• Acquire the BIUREV/BIUREV-N datasets.
• Python 3.
• barbar package (pip install barbar).
• wandb (pip install wandb).

Documentation

• data - a directory that handles data preparation and the PyTorch Dataset object.
• networks - a directory where the DSS and Ernst et al. models are defined.
• train.py - code for training a network.
• test.py - code for evaluating a network.
• losses.py - implements the loss functions.
• taskfiles - text files comprising lists of WAV files used for evaluation (in essence, all test files in BIUREV/BIUREV-N datasets).

Usage

1. Dataset Preparation

• Extract spectrograms from the BIUREV/BIUREV-N datasets to be used for training and validation.
  In data/prepare_dataset.py, set the variables clean_wavs_dir and reverb_wavs_dir as the directories where the clean and reverberant speech signals are stored. Makes sure to change the variables for both training(lines 154-155) and evaluation (lines 168-169) parts.

• Run:

  python data/prepare_dataset.py --mics_num <desired number of microphones> --dataset <BIUREV or BIUREV-N>
  

  This can take from a few minutes to a couple of hours, depending on the number of microphones.

2. Training

• Run:

python train.py --mics_num <choose from 1-8> --dataset <BIUREV or BIUREV-N> --unet_arch <dss or vanilla> --gpu_ids <for example 0 or 0,1,2,3>

This will create a new directory, named trained_networks.

3. Evaluation:

• Change line 208 such that the variable wavs_dir refers to the parent directory of BIUREV/BIUREV-N.
• Run:

python test.py --version_name <name of version> --dataset <BIUREV or BIUREV-N> -unet_arch <dss or vanilla>.

For example:

python test.py --version_name mics8_24.03.2021_05:33:29 --dataset BIUREV -unet_arch dss.

The number of microphones is inferred from the version's name.