Presentation 2021-03-04
An Estimated Intersections Reduction Method for Percussion Source Separation Based on the U-Net
Daisuke Tanaka, Susumu Kuroyanagi,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In the music information processing using drum information, the sound source separation is pre-processed to separate only drums from the music. The U-Net based sound source separation method has the highest separation accuracy among the conventional separation methods. However, it has a problem that the excess component tends to be generated at the intersection of the instruments, and the problem is critical in the drum separation where there are many intersections. In this report, in order to solve this problem by removing the excess components, we propose a method of separating the harmonic instrument sound at the intersections from the music and subtracting them from the drum separation result, and show the effectiveness of the method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Drums / Sound Source Separation / U-Net
Paper # NC2020-55
Date of Issue 2021-02-24 (NC)

Conference Information
Committee NC / MBE
Conference Date 2021/3/3(3days)
Place (in Japanese) (See Japanese page)
Place (in English) Online
Topics (in Japanese) (See Japanese page)
Topics (in English) Neuro Computing, Medical Engineering, etc.
Chair Kazuyuki Samejima(Tamagawa Univ) / Takashi Watanabe(Tohoku Univ.)
Vice Chair Rieko Osu(Waseda Univ.) / Ryuhei Okuno(Setsunan Univ.)
Secretary Rieko Osu(NTT) / Ryuhei Okuno(ATR)
Assistant Ken Takiyama(TUAT) / Nobuhiko Wagatsuma(Toho Univ.) / Akihiro Karashima(Tohoku Inst. of Tech.) / Jun Akazawa(Meiji Univ. of Integrative Medicine)

Paper Information
Registration To Technical Committee on Neurocomputing / Technical Committee on ME and Bio Cybernetics
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) An Estimated Intersections Reduction Method for Percussion Source Separation Based on the U-Net
Sub Title (in English)
Keyword(1) Drums
Keyword(2) Sound Source Separation
Keyword(3) U-Net
1st Author's Name Daisuke Tanaka
1st Author's Affiliation Nagoya Institute of Technology(NIT)
2nd Author's Name Susumu Kuroyanagi
2nd Author's Affiliation Nagoya Institute of Technology(NIT)
Date 2021-03-04
Paper # NC2020-55
Volume (vol) vol.120
Number (no) NC-403
Page pp.pp.71-76(NC),
#Pages 6
Date of Issue 2021-02-24 (NC)