Presentation 2005-11-25
Abnormality Extraction by Utilizing Video and Audio Sensors toward Human Action Recognition
Takashi HATTORI, Yoshinari KAMEDA, Yuichi OHTA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A stable and robust human action recognition is necessary for realization of high-value added living environments. Our purpose is to develop a sensing system for human action recognition by utilizing a number of multimodal sensors. In this paper, we describe a method to extract significant data segments from continuous video and audio data that include noise of daily lives. This method extracts the significant data segments by adaptive background model that can deal with stationary and ambient noise. We also discuss criteria to determine resolution of feature vector units so that they can be easily utilized for human action recognition in sensor fusion approach. We apply this method to some basic actions in an experimental environment. Finally, we also discuss adequate definition of similarity between the time-series data that are extracted by our method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Sensor fusion / stationary noise / action recognition / adaptive background model
Paper # MVE2005-51
Date of Issue

Conference Information
Committee MVE
Conference Date 2005/11/17(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 Media Experience and Virtual Environment (MVE)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Abnormality Extraction by Utilizing Video and Audio Sensors toward Human Action Recognition
Sub Title (in English)
Keyword(1) Sensor fusion
Keyword(2) stationary noise
Keyword(3) action recognition
Keyword(4) adaptive background model
1st Author's Name Takashi HATTORI
1st Author's Affiliation Graduate School of Systems and Information Engineering, University of Tsukuba()
2nd Author's Name Yoshinari KAMEDA
2nd Author's Affiliation Graduate School of Systems and Information Engineering, University of Tsukuba
3rd Author's Name Yuichi OHTA
3rd Author's Affiliation Graduate School of Systems and Information Engineering, University of Tsukuba
Date 2005-11-25
Paper # MVE2005-51
Volume (vol) vol.105
Number (no) 433
Page pp.pp.-
#Pages 6
Date of Issue