Anomaly Detection In Surveillance Videos

The goal of anomaly detection in surveillance videos is to promptly identify activities that deviate from normal patterns and determine the time windows during which anomalies occur. This task falls under the category of coarse-grained video understanding, using intelligent computer vision algorithms to automatically filter out abnormal events, thereby reducing the waste of human resources and time, and enhancing the intelligence level and efficiency of video surveillance systems.