Fig. 3From: Deep anomaly detection through visual attention in surveillance videosStep-by-step of finding visual attention regionBack to article page