Recognition of Abnormal Events in Surveillance Videos using Weakly Supervised Dual-Encoder Models
By: Noam Tsfaty , Avishai Weizman , Liav Cohen and more
Potential Business Impact:
Finds unusual things in videos automatically.
We address the challenge of detecting rare and diverse anomalies in surveillance videos using only video-level supervision. Our dual-backbone framework combines convolutional and transformer representations through top-k pooling, achieving 90.7% area under the curve (AUC) on the UCF-Crime dataset.
Similar Papers
Learning to Tell Apart: Weakly Supervised Video Anomaly Detection via Disentangled Semantic Alignment
CV and Pattern Recognition
Finds unusual events in videos better.
Transformer-Driven Multimodal Fusion for Explainable Suspiciousness Estimation in Visual Surveillance
CV and Pattern Recognition
Spots danger in crowds before it happens.
Pseudo Anomalies Are All You Need: Diffusion-Based Generation for Weakly-Supervised Video Anomaly Detection
CV and Pattern Recognition
Teaches computers to spot trouble without real examples.