Detection of Leakage Sound with Independent Component Analysis

• KOTANI Manabu

  Kobe University

• ARIMOTO Takahiko

  Kobe University

• OZAWA Seiichi

  Kobe University

• AKAZAWA Kenzo

  Osaka University

Bibliographic Information

Other Title

• 458 独立成分分析による漏洩音の検出

Abstract

We describe what characteristics an independent component analysis can extract from leakage sounds. The learning algorithm of a network was an information-maximization approach proposed by Bell and Sejnowski. We applied the independent component analysis to the leakage sound. After the learning, most of the basis functions that are columns of a mixing matrix were localized in frequency. Furthermore, there were some basis functions to extract the feature of leakage sound. From these results, we confirmed that the application of independent component analysis to the signal processing of the acoustic signal was effective.

Journal

• Dynamics and Design Conference : 機械力学・計測制御講演論文集 : D & D

  Dynamics and Design Conference : 機械力学・計測制御講演論文集 : D & D 2000 (0), 236-, 2000-09-01

  The Japan Society of Mechanical Engineers

Keywords

• Independent Component Analysis
• Leakage Sound
• Neural Networks

Details

• CRID: 1573950401935102720
• NII Article ID: 110002482598
• NII Book ID: AA11901770
• ISSN: 13480235
• Text Lang: ja
• Data Source:
  • CiNii Articles