CrossNet-Open-Unmix for Music Source Separation (X-UMXL)

Ryosuke Sawata, Stefan Uhlich, Shusuke Takahashi, Yuki MitsufujiZenodo 2022 · DOI · OpenAlex

Abstract

Weights of CrossNet-Open-Unmix (X-UMX) trained on the internal 100h dataset which is larger than MUSDB18, named X-UMX Large (X-UMXL). The weights can be used with X-UMX on Asteroid (PyTorch). The details of X-UMX are described in here.
BibTeX
@inproceedings{sawata2022crossnet,
  author = {Ryosuke Sawata and Stefan Uhlich and Shusuke Takahashi and Yuki Mitsufuji},
  title = {{CrossNet-Open-Unmix for Music Source Separation (X-UMXL)}},
  booktitle = {Zenodo},
  year = {2022},
  doi = {10.5281/zenodo.7128658},
}

Loading related content...

Figure