Figure 9From: Robust Recognition of Specific Human Behaviors in Crowded Surveillance Video Sequences Detecting candidate human regions. Back to article page