Fig. 5From: Deep anomaly detection through visual attention in surveillance videosThe illustration of different block size used in the proposed texture descriptorBack to article page