Music source separation in the waveform domain

Abstract: Source separation for music is the task of isolating contributions, or stems, from different instruments recorded individually and arranged together to form a song. Such components include voice, bass, drums and any other accompaniments. Contrarily to many audio synthesis tasks where the best performances are achieved by models that directly generate the waveform, the state-of-the-art in source separation for music is to compute masks on the magnitude spectrum. In this paper, we compare two waveform domain architectures. We first adapt Conv-Tasnet, initially developed for speech source separation, to the task of music source separation. While ConvTasnet beats many existing spectrogram-domain methods, it suffers from significant artifacts, as shown by human evaluations. We propose instead Demucs, a novel waveform-to-waveform model, with a U-Net structure and bidirectional LSTM. Experiments on the MusDB dataset show that, with proper data augmentation, Demucs beats all existing state-of-the-art architectures, including Conv-Tasnet, with 6.3 SDR on average, (and up to 6.8 with 150 extra training songs, even surpassing the IRM oracle for the bass source). Using recent development in model quantization, Demucs can be compressed down to 120MB without any loss of accuracy. We also provide human evaluations, showing that Demucs benefit from a large advantage in terms of the naturalness of the audio. However, it suffers from some bleeding, especially between the vocals and other source.

Alexandre Défossez, Nicolas Usunier, Léon Bottou and Francis Bach: Music source separation in the waveform domain, arXiv preprint arXiv:1911.13254, 2019.

defossez-2019.djvu defossez-2019.pdf defossez-2019.ps.gz

@article{defossez-2019,
  title = {Music source separation in the waveform domain},
  author = {D{\'e}fossez, Alexandre and Usunier, Nicolas and Bottou, L{\'e}on and Bach, Francis},
  journal = {arXiv preprint arXiv:1911.13254},
  year = {2019},
  url = {http://leon.bottou.org/papers/defossez-2019},
}