Binaural Sound Source Distance Estimation and Localization for a Moving Listener

This is a repository implementing the methods described in the following paper:

If you use this repository or the dataset associated with it, kindly consider citing the abovementioned paper.

DATASETS

This repository consists of multiple Python scripts described below.

The batch_feature_extraction.py is a standalone wrapper script, that extracts the features, labels, and normalizes the training and test split features for a given dataset. Make sure you update the location of the downloaded datasets before.
The parameters.py script consists of all the training, model, and feature parameters. If a user has to change some parameters, they have to create a sub-task with unique id here. Check code for examples.
The cls_feature_class.py script has routines for labels creation, features extraction and normalization.
The cls_data_generator.py script provides feature + label data in generator mode for training.
The models.py script implements the different model architectures related to the paper mentioned above.
The cls_metric.py script implements the metrics for localization and distance estimation.
The rnd_files.npy is a file containing an exemplary random shuffle of the dataset indices (used for repeatability between experiments). Can changed for a different one.
The train_doa_sde_joint.py is a wrapper script that trains SDEL models with a single task approach.
The train_doa_sde_joint_multi.py is a wrapper script that trains SDEL models with a multi task approach.

The provided codebase has been tested on Python 3.9 and Torch 1.13

This repository is an adapted version of doa-net. The repository is licensed under the TAU License.