DMCA
Observe locally, infer globally: A space-time mrf for detecting abnormal activities with incremental updates
Cached
Download Links
| Venue: | In CVPR |
| Citations: | 42 - 0 self |
BibTeX
@INPROCEEDINGS{Kim_observelocally,,
author = {Jaechul Kim and Kristen Grauman},
title = {Observe locally, infer globally: A space-time mrf for detecting abnormal activities with incremental updates},
booktitle = {In CVPR},
year = {},
pages = {3318}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We propose a space-time Markov Random Field (MRF) model to detect abnormal activities in video. The nodes in the MRF graph correspond to a grid of local regions in the video frames, and neighboring nodes in both space and time are associated with links. To learn normal patterns of activity at each local node, we capture the distribution of its typical optical flow with a Mixture of Probabilistic Principal Component Analyzers. For any new optical flow patterns detected in incoming video clips, we use the learned model and MRF graph to compute a maximum a posteriori estimate of the degree of normality at each local node. Further, we show how to incrementally update the current model parameters as new video observations stream in, so that the model can efficiently adapt to visual context changes over a long period of time. Experimental results on surveillance videos show that our space-time MRF model robustly detects abnormal activities both in a local and global sense: not only does it accurately localize the atomic abnormal activities in a crowded video, but at the same time it captures the global-level abnormalities caused by irregular interactions between local activities. 1.
Keyphrases
abnormal activity space-time mrf incremental update local node irregular interaction new video observation stream surveillance video show local activity new optical flow pattern video frame crowded video local region probabilistic principal component analyzer atomic abnormal activity current model parameter space-time markov random field normal pattern long period space-time mrf model video clip typical optical flow visual context change global-level abnormality learned model mrf graph correspond posteriori estimate mrf graph experimental result global sense
