Musical Part Separation Based on Perceptual Hierarchy

  • KINOSHITA Tomoyoshi
    Graduate School of Engineering, The University of Tokyo
  • HANDA Ibuki
    Graduate School of Engineering, The University of Tokyo
  • MUTO Makoto
    Graduate School of Engineering, The University of Tokyo
  • SAKAI Shuichi
    Graduate School of Information Science and Technology, The University of Tokyo
  • TANAKA Hidehiko
    Graduate School of Information Science and Technology, The University of Tokyo

Search this article

Description

Previously, there were a number of works on auditory scene analysis. Especially in the field of music recognition, several works on automatic transcription were done. However, conventional method used only the temporarily local processes and could not achieve practical performance. In this paper, we proposed a novel method to improve this problem. The proposed method focused on the perceptual hierarchy of music stream; notes, phrases and parts. We used local clues to extract phrases from notes, and used global clues to extract parts from phrases. This hierarchical process enabled to avoid computational explosin. In experimental result, the system succeeded to extract parts in about 80-90% accuracy.

Journal

Keywords

Details 詳細情報について

  • CRID
    1573387452279393536
  • NII Article ID
    110003219957
  • NII Book ID
    AA10826272
  • ISSN
    09168532
  • Text Lang
    en
  • Data Source
    • CiNii Articles

Report a problem

Back to top