Music Signal Separation by Orthogonality and Maximum-Distance Constrained Nonnegative Matrix Factorization with Target Signal Information

Yagi K, Takahashi Y, Saruwatari H, Shikano K, Kondo K. Music Signal Separation by Orthogonality and Maximum-Distance Constrained Nonnegative Matrix Factorization with Target Signal Information. In: AES Conference: 45th International Conference: Applications of Time-Frequency Processing in Audio. Audio Engineering Society; 2012. Paper 2-5. Available from: https://aes.org/publications/elibrary-page/?id=16207

@inproceedings{Yagi2012_16207,
  author = {Yagi, Kosuke and Takahashi, Yu and Saruwatari, Hiroshi and Shikano, Kiyohiro and Kondo, Kazunobu},
  title = {{Music Signal Separation by Orthogonality and Maximum-Distance Constrained Nonnegative Matrix Factorization with Target Signal Information}},
  booktitle = {AES Conference: 45th International Conference: Applications of Time-Frequency Processing in Audio},
  note = {Paper 2-5},
  year = {2012},
  month = mar,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=16207}
}

TY  - CPAPER
TI  - Music Signal Separation by Orthogonality and Maximum-Distance Constrained Nonnegative Matrix Factorization with Target Signal Information
AU  - Yagi, Kosuke
AU  - Takahashi, Yu
AU  - Saruwatari, Hiroshi
AU  - Shikano, Kiyohiro
AU  - Kondo, Kazunobu
T2  - AES Conference: 45th International Conference: Applications of Time-Frequency Processing in Audio
M1  - Paper 2-5
PY  - 2012
DA  - 2012/03/06
UR  - https://aes.org/publications/elibrary-page/?id=16207
PB  - Audio Engineering Society
LA  - en
AB  - In this paper, we address the separation of multiple instrumental sources based on semi-supervised nonnegative matrix factorization (SNMF) and propose a new constrained SNMF. Recently, various types of SNMF have been proposed. In particular, we focus our attention on one type of SNMF that utilizes information on a priori bases. Indeed, this type of SNMF can achieve better separation performance. However, SNMF without are any constraint between a priori bases and other bases often degrades separation performance. Thus, we propose a new SNMF that imposes a constraint between a priori bases and other bases. An experimental result shows the efficacy of the proposed constrained SNMF.
ER  -