Source Separation Landing Page

September 12, 2025 · View on GitHub

Please consider giving back to the community if you have benefited from these works

If you've benefited commercially from these works, which we've poured significant effort into and released under permissive licenses, we hope you've found it valuable! While these licenses give you lots of freedom, we believe in nurturing a vibrant ecosystem where innovation can continue to flourish.

So, as a gesture of appreciation and responsibility, we strongly urge commercial entities that have gained from this software to consider making voluntary contributions to music-related non-profit organizations of your choice. Your contribution directly helps support the foundational work that empowers your commercial success and ensures open-source innovation keeps moving forward.

Some suggestions for the beneficiaries are provided here. Please do not hesitate to contribute to the list by opening pull requests there.

Source Separation Landing Page

Note: all the model repositories are currently being reworked into one unified repository to make it easier to maintain. The alpha version is here.

Bandit: Cinematic Audio Source Separation

K. N. Watcharasupat et al., “A Generalized Bandsplit Neural Network for Cinematic Audio Source separation,” IEEE Open Journal of Signal Processing, vol. 5, pp. 73–81, 2023, doi: 10.1109/OJSP.2023.3339428.

Bandit is a cinematic audio source separation model adapted from Bandsplit RNN.

Banquet: Query-based Music Source Separation

K. N. Watcharasupat and A. Lerch, “A Stem-Agnostic Single-Decoder System for Music Source Separation Beyond Four Stems,” to appear in the Proceedings of the 25th International Society for Music Information Retrieval Conference, San Francisco, CA, USA, Nov. 2024.

Banquet is a query-based music source separation model adapted from Bandit + PaSST

Banquet for Cinematic Audio Source Separation

K. N. Watcharasupat, C.-W. Wu, and I. Orife, “Facing the Music: Tackling Singing Voice Separation in Cinematic Audio Source Separation,” to appear in the Late-Breaking Demo Session of the 25th International Society for Music Information Retrieval Conference, San Francisco, CA, USA, Nov. 2024.

Divide and Remaster v3: Multilingual Cinematic Audio Source Separation Dataset

K. N. Watcharasupat, C.-W. Wu, and I. Orife, “Remastering Divide and Remaster: A Cinematic Audio Source Separation Dataset with Multilingual Support,” in Proceedings of the 5th IEEE International Symposium on the Internet of Sounds, Erlangen, Germany: IEEE, Sep. 2024.

A multilingual rework of the Divide and Remaster v2 dataset.

SPAuQ: Spatial Decomposition for SNR-like evaluation

K. N. Watcharasupat and A. Lerch, “Quantifying Spatial Audio Quality Impairment,” in Proceedings of the 2024 International Conference on Acoustics, Speech, and Signal Processing, Seoul, Korea, Republic of: IEEE, Apr. 2024, pp. 746–750. doi: 10.1109/ICASSP48485.2024.10447947.