Structure Learning and Source Separation of Music Audio Signals based on Bayesian Nonparametrics

2012 Fiscal Year Final Research Report

• Project/Area Number: 23700184
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Intelligent informatics
• Research Institution: National Institute of Advanced Industrial Science and Technology
• Principal Investigator: YOSHII Kazuyoshi 独立行政法人産業技術総合研究所, 情報技術研究部門, 主任研究員 (20510001)
• Project Period (FY): 2011 – 2012
• Keywords: 音楽情報処理 / 機械学習

Research Abstract

This research aims to develop a nonparametric Bayesian model that can be used for jointly performing source separation and structure learning of music audio signals in an unsupervised manner. First, we have studied on structure learning of music scores (s

Research Products

• [Presentation] Infinite Positive Semidefinite Tensor Factorization for Source Separation of Mixture Signals (2013)
  • Author(s): Kazuyoshi Yoshii, Ryota Tomioka,Daichi Mochihashi, and Masataka Goto
  • Organizer: The 30th International Conference on Machine Learning (ICML)
  • Place of Presentation: Atlanta (USA)
  • Year and Date: 2013-06-18
• [Presentation] Infinite Kernel Linear Prediction for Joint Estimation of Spectral Envelope and Fundamental Frequency (2013)
  • Author(s): Kazuyoshi Yoshii and Masataka Goto
  • Organizer: IEEE International Conference on Acoustics,Speech, and Signal Processing (ICASSP)
  • Place of Presentation: Vancouver (Canada)
  • Year and Date: 2013-05-29
• [Presentation] スペクトル包絡と基本周波数の同時推定のための無限カーネル線形予測分析法 (2013)
  • Author(s): 吉井和佳, 後藤真孝
  • Organizer: 情報処理学会 第99回音楽情報科学研究会
  • Place of Presentation: お茶の水女子大学(東京都)
  • Year and Date: 2013-05-11
• [Presentation] 音楽音響信号解析のためのガンマ過程に基づく無限複合自己回帰モデル (2012)
  • Author(s): 吉井和佳, 後藤真孝
  • Organizer: 第15回情報論的学習理論ワークショップ, IBISML2012-51
  • Place of Presentation: 筑波大学東京キャンパス(東京都)
  • Year and Date: 2012-11-07
• [Presentation] Infinite Composite Autoregressive Models for Music Signal Analysis (2012)
  • Author(s): Kazuyoshi Yoshii and Masataka Goto
  • Organizer: The 13th International Society for Music Information Retrieval Conference (ISMIR)
  • Place of Presentation: Porto (Portugal)
  • Year and Date: 2012-10-09
• [Presentation] 多重音基本周波数解析のための無限複合自己回帰モデル (2012)
  • Author(s): 吉井和佳, 後藤真孝
  • Organizer: 情報処理学会 第96回音楽情報科学研究会
  • Place of Presentation: 近江町交流プラザ(石川県)
  • Year and Date: 2012-08-10
• [Presentation] Unsupervised Music Understanding Based on Nonparametric Bayesian Models (2012)
  • Author(s): Kazuyoshi Yoshii and Masataka Goto
  • Organizer: IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)
  • Place of Presentation: Kyoto International Conference Center (Kyoto)
  • Year and Date: 2012-03-02
• [Presentation] A Unified Probabilistic Model of Note Combinations and Chord Progressions (2011)
  • Author(s): Kazuyoshi Yoshii and Masataka Goto
  • Organizer: The 4th International Workshop on Music and Machine Learning (MML)
  • Place of Presentation: Granada (Spain)
  • Year and Date: 2011-12-17
• [Presentation] A Vocabulary-Free Infinity-Gram Model for Nonparametric Bayesian Chord Progression Analysis (2011)
  • Author(s): Kazuyoshi Yoshii and Masataka Goto
  • Organizer: The 12th International Society for Music Information Retrieval Conference (ISMIR)
  • Place of Presentation: Miami (USA)
  • Year and Date: 2011-10-27
• [Presentation] 和音進行解析のための語彙フリー無限グラムモデル (2011)
  • Author(s): 吉井和佳, 後藤真孝
  • Organizer: 情報処理学会 第91回音楽情報科学研究会
  • Place of Presentation: 関西学院大学(兵庫県)
  • Year and Date: 2011-07-27
• [Remarks]
  • URL: http://staff.aist.go.jp/k.yoshii/