Presentation 2010-10-08
An abnormal sound detection method using multi-stage GMM for surveillance microphone
Akinori ITO, Akihito AIBA, Masashi ITO, Shozo MAKINO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We have been developing a method for detecting abnormal sound events from audio signal recorded in real environments, which uses the multi-stage Gaussian Mixture Models (GMM) that learns rare sounds using multiple GMMs. In this paper, we investigate relationship between sound environment and detection performance, and we found that the performance deteriorates in noisy environments. The performance largely depended on SN ratio of the abnormal sounds. Next, we investigated methods for determining hyperparameters of the multi-stage GMM, which involves intermediate thresholds, numbers of mixture of GMMs and the detection threshold. From the experimental results, combination of Percentile-based threshold determination and Bayesian information criterion (BIC)-based mixture determination was most effective. However, when using the automatically-determined parameters, the detection performance deteriorated around 20%.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) CCTV / incident detection / abnormal sound detection / Gaussian mixture model
Paper # PRMU2010-88,SP2010-44,WIT2010-32
Date of Issue

Conference Information
Committee WIT
Conference Date 2010/10/1(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Well-being Information Technology(WIT)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) An abnormal sound detection method using multi-stage GMM for surveillance microphone
Sub Title (in English)
Keyword(1) CCTV
Keyword(2) incident detection
Keyword(3) abnormal sound detection
Keyword(4) Gaussian mixture model
1st Author's Name Akinori ITO
1st Author's Affiliation Graduate School of Engineering, Tohoku University()
2nd Author's Name Akihito AIBA
2nd Author's Affiliation Graduate School of Engineering, Tohoku University
3rd Author's Name Masashi ITO
3rd Author's Affiliation Tohoku Institute of Technology
4th Author's Name Shozo MAKINO
4th Author's Affiliation Tohoku Bunka Gakuen University
Date 2010-10-08
Paper # PRMU2010-88,SP2010-44,WIT2010-32
Volume (vol) vol.110
Number (no) 221
Page pp.pp.-
#Pages 6
Date of Issue