Title :
Memory-Based Attention Control for Activity Recognition at a Subway Station
Author :
MacDorman, Karl F. ; Nobuta, Hiroshi ; Koizumi, Satoshi ; Ishiguro, Hiroshi
Author_Institution :
Indiana Univ.
Abstract :
We have developed a multicamera system, Digital City Surveillance, which uses a new calibration-free behavior recognition method for monitoring human activity at a subway station. We trained nine support vector machines from operator-classified data to recognize 512 combinations of events. Our method of attention control greatly reduced computation and increased classification accuracy
Keywords :
cameras; image recognition; support vector machines; video surveillance; Digital City Surveillance; activity recognition; calibration-free behavior recognition; human activity monitoring; memory-based attention control; multicamera system; subway station; support vector machine; Cameras; Cities and towns; Computer vision; Control systems; Humans; Monitoring; Sensor systems; Support vector machine classification; Support vector machines; Surveillance; distributed vision; feature extraction; human activity recognition; omnidirectional; support vector machines; video surveillance;
Journal_Title :
MultiMedia, IEEE
DOI :
10.1109/MMUL.2007.39
