English

KUIELab-MDX-Net: A Two-Stream Neural Network for Music Demixing

Audio and Speech Processing 2021-11-25 v1 Sound

Abstract

Recently, many methods based on deep learning have been proposed for music source separation. Some state-of-the-art methods have shown that stacking many layers with many skip connections improve the SDR performance. Although such a deep and complex architecture shows outstanding performance, it usually requires numerous computing resources and time for training and evaluation. This paper proposes a two-stream neural network for music demixing, called KUIELab-MDX-Net, which shows a good balance of performance and required resources. The proposed model has a time-frequency branch and a time-domain branch, where each branch separates stems, respectively. It blends results from two streams to generate the final estimation. KUIELab-MDX-Net took second place on leaderboard A and third place on leaderboard B in the Music Demixing Challenge at ISMIR 2021. This paper also summarizes experimental results on another benchmark, MUSDB18. Our source code is available online.

Keywords

Cite

@article{arxiv.2111.12203,
  title  = {KUIELab-MDX-Net: A Two-Stream Neural Network for Music Demixing},
  author = {Minseok Kim and Woosung Choi and Jaehwa Chung and Daewon Lee and Soonyoung Jung},
  journal= {arXiv preprint arXiv:2111.12203},
  year   = {2021}
}

Comments

MDX Workshop @ ISMIR 2021, 7 pages, 3 figures