Explainable AI for Suspicious Activity Detection in Surveillance

This research explores the application of Transformer models to fuse multimodal data for improved suspicious activity detection in visual surveillance. The emphasis on explainability is crucial for building trust and enabling practical application in security contexts.