Modal Expansion-based Data Generation Approach for Deep Learning-Enabled Sound Source Localization in a Small Enclosure

This repository contains the python implementation for the dataset generation part of the paper "Modal Expansion-based Data Generation Approach for Deep Learning-Enabled Sound Source Localization in a Small Enclosure".

Dataset

• Source signals: LibriSpeech

These datasets mentioned above can be downloaded from this OneDrive link.

The data directory structure is shown as follows:

.
|---data
    |---LibriSpeech
        |---dev-clean
        |---test-clean
        |---train-clean-100
    |---test (generated)
    |---train (generated)
    |---dev (generated)

Get Started

Dependencies

We strongly recommend that you can use VSCode and Docker for this project, it can save you much time😁! Note that the related configurations has already been within .devcontainer. The detail information can be found in this Tutorial_for_Vscode&Dokcer.

Configurations

The realted configurations are all saved in config/.

• The dataSIMU.yaml is used to configure the data generation.

You can change the value of these items based on your need.

Note: Do not forget to install webrtcvad.

🚀 Quick Start

• Data Generation

Generate the training/val/test data:

bash scripts/datasimu.sh

🎓 Citation

If you find our work useful in your research, please consider citing:

@article{pi2026modal,
  title={Modal expansion-based data generation approach for deep learning-enabled sound source localization in a small enclosure},
  author={Pi, Rendong and Yu, Xiang},
  journal={Applied Acoustics},
  volume={241},
  pages={111023},
  year={2026},
  publisher={Elsevier}
}