Minimum-entropy models of scene activity
Kettnaker, V.
Brand, M.
Dept. of Comput. Sci., Cornell Univ., Ithaca, NY ;
This paper appears in: Computer Vision and Pattern Recognition, 1999. IEEE Computer Society Conference on.
Publication Date: 1999
Volume: 1,
On page(s): -286 Vol. 1
Meeting Date: 06/23/1999 - 06/25/1999
Location: Fort Collins, CO, USA
ISBN: 0-7695-0149-4
References Cited: 15
INSPEC Accession Number: 6331105
Digital Object Identifier: 10.1109/CVPR.1999.786952
Current Version Published: 2002-08-06
Abstract
We show how to learn a concise, interpretable model of scene
activity directly from optical flow. The model represents the principal
routes and modes of movement in complex scenes such as pedestrian plazas
and traffic intersections, and supports a variety of inferences about
the observed activities, including annotation, prediction, and anomaly
detection. The model takes the form of a novel hidden Markov model
generalization that observes a variable number of datapoints per frame
(time step). A monotonic entropy-optimizing algorithm determines the
parameters and structure of this model, exploiting the duality between
learning and compression to produce highly predictive and interpretable
models. This approach discovers minimal models of coherent motions and
their switching dynamics-without tracking or prior knowledge about the
spatial or temporal structure of the scene
Index
Terms
Available to subscribers and IEEE members.
References
Available to subscribers and IEEE members.
Citing Documents
Available to subscribers and IEEE members.
Cart
