Recognition of Abnormal Events in Surveillance Videos using Weakly Supervised Dual-Encoder Models
Preprint
Abstract
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.Keywords
All Related Versions
This publication has 3 references indexed in Scilit:
- MSTAgent-VAD: Multi-scale video anomaly detection using time agent mechanism for segments’ temporal context miningExpert Systems with Applications, 2025
- Weakly Supervised Video Anomaly Detection and Localization with Spatio-Temporal Prompts2024
- Unsupervised video anomaly detection via normalizing flows with implicit latent featuresPattern Recognition, 2022