#Crowd Surveillance

Behavioral Anomaly Detection in High-Density Public Environments: Spatio-Temporal Graph Convolutional Networks (ST-GCN) and Optical Flow Modeling

A scientific study on automated behavioral anomaly detection in crowded environments using Spatio-Temporal Graph Convolutional Networks (ST-GCN), dual-stream optical flow, variational autoencoders, and edge crowd density dynamics.