TY - CONF
T1 - Activity recognition using the dynamics of the configuration of interacting objects
T2 - Computer Vision and Pattern Recognition, 2003. Proceedings. 2003 IEEE Computer Society Conference on
Y1 - 2003
A1 - Vaswani, N.
A1 - RoyChowdhury, A.
A1 - Chellapa, Rama
KW - 2D
KW - abnormal
KW - abnormality
KW - abnormality;
KW - acoustic
KW - activity
KW - analysis;
KW - change;
KW - Computer
KW - configuration
KW - configuration;
KW - data;
KW - DETECTION
KW - detection;
KW - distribution;
KW - drastic
KW - dynamics;
KW - event;
KW - filter;
KW - hand-picked
KW - image
KW - infrared
KW - interacting
KW - learning;
KW - location
KW - low
KW - mean
KW - model;
KW - monitoring;
KW - MOTION
KW - moving
KW - noise;
KW - noisy
KW - object
KW - object;
KW - observation
KW - observation;
KW - particle
KW - pattern
KW - plane;
KW - point
KW - polygonal
KW - probability
KW - probability;
KW - problem;
KW - processing;
KW - radar
KW - recognition;
KW - resolution
KW - sensor;
KW - sensors;
KW - sequence;
KW - SHAPE
KW - shape;
KW - signal
KW - slow
KW - statistic;
KW - strategy;
KW - Surveillance
KW - surveillance;
KW - target
KW - test
KW - tracking;
KW - video
KW - video;
KW - visible
KW - vision;
AB - Monitoring activities using video data is an important surveillance problem. A special scenario is to learn the pattern of normal activities and detect abnormal events from a very low resolution video where the moving objects are small enough to be modeled as point objects in a 2D plane. Instead of tracking each point separately, we propose to model an activity by the polygonal 'shape' of the configuration of these point masses at any time t, and its deformation over time. We learn the mean shape and the dynamics of the shape change using hand-picked location data (no observation noise) and define an abnormality detection statistic for the simple case of a test sequence with negligible observation noise. For the more practical case where observation (point locations) noise is large and cannot be ignored, we use a particle filter to estimate the probability distribution of the shape given the noisy observations up to the current time. Abnormality detection in this case is formulated as a change detection problem. We propose a detection strategy that can detect both 'drastic' and 'slow' abnormalities. Our framework can be directly applied for object location data obtained using any type of sensors - visible, radar, infrared or acoustic.
JA - Computer Vision and Pattern Recognition, 2003. Proceedings. 2003 IEEE Computer Society Conference on
VL - 2
M3 - 10.1109/CVPR.2003.1211526
ER -