Back close

Object-centric and memory-guided network-based normality modeling for video anomaly detection

Publication Type : Journal Article

Publisher : Springer

Source : Signal,Image and Video Processing,pp.1-7

Url :

Campus : Chennai

School : School of Computing

Department : Computer Science and Engineering

Year : 2022

Abstract : Anomaly detection in surveillance videos is a challenging and demanding task. Autoencoders trained on segments of normal events are expected to give high reconstruction error for abnormal events than that for normal events. However, the assumption of autoencoders giving high reconstruction error is not always true in practice. Since the autoencoder sometimes offers better generalization, it also reconstructs abnormal events well, leading to slightly degraded performance for anomaly detection. Another issue is that the performance of real-time anomalous activity detection in surveillance videos still needs improvement. To address these issues, we propose an Object-centric and Memory-guided residual spatiotemporal autoencoder (OM-RSTAE) to detect video anomalies. The proposed technique achieved improved results over benchmark datasets, namely UCSD-Ped2, Avenue, ShanghaiTech and UCF-Crime datasets.

Cite this Research Publication : Chandrakala,S.,Shalmiya,P.,Srinivas,V.andDeepak,K.,2022.Object-centric and memory-guided network-based normality modeling for video anomaly detection. Signal,Image and Video Processing,pp.1-7.

Admissions Apply Now