Software & Data Downloads — hyper-unmix

Hyperbolic Audio Source Separation for training and interacting with models proposed in the ICASSP 2023 paper, "Hyperbolic Audio Source Separation".

PyTorch implementation for training and interacting with models proposed in our ICASSP 2023 paper, “Hyperbolic Audio Source Separation.” We include the weights for a model pre-trained on the Librispeech Slakh Unmix (LSX) dataset, which hierarchically separates an audio mixture containing music and speech. Furthermore, code for training models using mask cross-entropy, spectrogram, and waveform losses is included. An interface for interacting with the learned hyperbolic embeddings created using PyQT6 is also provided in this codebase.

  •  Petermann, D., Wichern, G., Subramanian, A.S., Le Roux, J., "Hyperbolic Audio Source Separation", IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), DOI: 10.1109/​ICASSP49357.2023.10094943, May 2023, pp. 1-5.
    BibTeX TR2023-017 PDF Video Software
    • @inproceedings{Petermann2023may,
    • author = {Petermann, Darius and Wichern, Gordon and Subramanian, Aswin Shanmugam and Le Roux, Jonathan},
    • title = {Hyperbolic Audio Source Separation},
    • booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)},
    • year = 2023,
    • pages = {1--5},
    • month = may,
    • publisher = {IEEE},
    • doi = {10.1109/ICASSP49357.2023.10094943},
    • url = {https://www.merl.com/publications/TR2023-017}
    • }

Access software at https://github.com/merlresearch/hyper-unmix.