PROJECT TITLE :

Abnormal Event Detection via Compact Low-Rank Sparse Learning

ABSTRACT:

Sparsity-based ways have been recently applied to abnormal event detection, and have achieved impressive results. But, most such methods fail to consider the link among coefficient vectors; furthermore, they neglect the underlying "dictionary structure."' The authors' compact low-rank sparse representation (CLSR) methodology overcomes these drawbacks. Specifically, it adds compact regularization to the sparse representation model, that explicitly considers the connection among coefficient vectors. The authors utilize the low-rank property to capture the underlying dictionary structure. Their methodology is verified on three difficult databases, and therefore the experimental results demonstrate that it compares favorably to state-of-the-art ways in abnormal event detection.