口ドラム認識手法とそのドラム譜入力システムへの応用  [in Japanese] A Voice Percussion Recognition Method and Its Application to a Music Notation System of Drum Sounds  [in Japanese]

Access this Article

Search this Article

Author(s)

    • 中野 倫靖 NAKANO TOMOYASU
    • 筑波大学大学院図書館情報メディア研究科 Graduate School of Library, Information and Media Studies, University of Tsukuba
    • 緒方 淳 OGATA JUN
    • 産業技術総合研究所 National Institute of Advanced Industrial Science and Technology (AIST)
    • 平賀 譲 HIRAGA YUZURU
    • 筑波大学大学院図書館情報メディア研究科 Graduate School of Library, Information and Media Studies, University of Tsukuba

Abstract

本論文では,人がドラムの音を「ドンタンドドタン」のように擬音語で真似た音声(口ドラム)を,それに対応するドラムパターンとして認識する手法を提案する.口ドラムには,声質などの発声スタイルの個人差と表現の個人差の2 種類の個人差が存在し,認識手法はこれらの個人差を吸収できるものが望ましい.従来,実際のドラム音(楽器音)を対象とした認識は研究されてきたが,それらの手法では口ドラムの多様な個人差への対処が難しかった.そこで本手法では,口ドラムを,その各音を表す音素列の擬音語で表現し,多様な擬音語表現の辞書を用意することで,表現の個人差に対処する.さらに,音声認識で用いられている音響モデルを用いることにより,各歌唱者間の声質の個人差を吸収する.本手法を200 発話の口ドラムデータに対して適用した結果,最も良い実験条件において92.0%の認識率を得た.この結果は,提案手法が十分実用性を持つことを示しており,応用例として,口ドラムによるドラム譜入力インタフェースVoice Drummer を実装した.This paper presents a method of recognizing voice percussion (verbalized expression of drum sound by voice) as an expression of intended drum patterns. Recognition of voice percussion requires an approach that is different from existing methods for drum sound recognition. Individual differences in both vocal characteristics and the verbal expressions used add further complication to the task. The approach taken in this study uses phonemic sequences of onomatopoeia as internal representation of drum sounds. The set of onomatopoeia used in drum sounds are included in a pronunciation dictionary, and the phonemic sequences are estimated by utilizing an acoustic model. The acoustic model and the dictionary are intended to deal with the two types of individual differences mentioned above. In a recognition experiment with 200 utterances of voice percussion, our method achieved a recognition rate of 92.0% for the highest-tuned setting. Following the results of the proposed method, Voice Drummer, a music notation interface of drum sounds, was implemented, as a practical application for voice percussion recognition.

This paper presents a method of recognizing voice percussion (verbalized expression of drum sound by voice) as an expression of intended drum patterns. Recognition of voice percussion requires an approach that is different from existing methods for drum sound recognition. Individual differences in both vocal characteristics and the verbal expressions used add further complication to the task. The approach taken in this study uses phonemic sequences of onomatopoeia as internal representation of drum sounds. The set of onomatopoeia used in drum sounds are included in a pronunciation dictionary, and the phonemic sequences are estimated by utilizing an acoustic model. The acoustic model and the dictionary are intended to deal with the two types of individual differences mentioned above. In a recognition experiment with 200 utterances of voice percussion, our method achieved a recognition rate of 92.0% for the highest-tuned setting. Following the results of the proposed method, Voice Drummer, a music notation interface of drum sounds, was implemented, as a practical application for voice percussion recognition.

Journal

  • IPSJ journal

    IPSJ journal 48(1), 386-397, 2007-01-15

    Information Processing Society of Japan (IPSJ)

References:  21

Cited by:  3

Codes

  • NII Article ID (NAID)
    110006152212
  • NII NACSIS-CAT ID (NCID)
    AN00116647
  • Text Lang
    JPN
  • Article Type
    Journal Article
  • ISSN
    1882-7764
  • NDL Article ID
    8650008
  • NDL Source Classification
    ZM13(科学技術--科学技術一般--データ処理・計算機)
  • NDL Call No.
    Z14-741
  • Data Source
    CJP  CJPref  NDL  NII-ELS  IR  IPSJ 
Page Top