Deep anomaly detection through visual attention in surveillance videos
- PDF / 2,358,534 Bytes
- 17 Pages / 595.276 x 790.866 pts Page_size
- 84 Downloads / 298 Views
Open Access
RESEARCH
Deep anomaly detection through visual attention in surveillance videos Nasaruddin Nasaruddin1,2, Kahlil Muchtar1,2,3* , Afdhal Afdhal1,2 and Alvin Prayuda Juniarta Dwiyantoro3 *Correspondence: [email protected] 1 Department of Electrical and Computer Engineering, Syiah Kuala University, Aceh, CO 23111, Indonesia Full list of author information is available at the end of the article
Abstract This paper describes a method for learning anomaly behavior in the video by finding an attention region from spatiotemporal information, in contrast to the full-frame learning. In our proposed method, a robust background subtraction (BG) for extracting motion, indicating the location of attention regions is employed. The resulting regions are finally fed into a three-dimensional Convolutional Neural Network (3D CNN). Specifically, by taking advantage of C3D (Convolution 3-dimensional), to completely exploit spatiotemporal relation, a deep convolution network is developed to distinguish normal and anomalous events. Our system is trained and tested against a large-scale UCFCrime anomaly dataset for validating its effectiveness. This dataset contains 1900 long and untrimmed real-world surveillance videos and splits into 950 anomaly events and 950 normal events, respectively. In total, there are approximately ~ 13 million frames are learned during the training and testing phase. As shown in the experiments section, in terms of accuracy, the proposed visual attention model can obtain 99.25 accuracies. From the industrial application point of view, the extraction of this attention region can assist the security officer on focusing on the corresponding anomaly region, instead of a wider, full-framed inspection. Keywords: Visual attention approach, Convolutional neural network (CNN), Integrated surveillance system, Anomaly classification
Introduction While monitoring of public violence for safety and security is becoming increasingly important, surveillance systems are now being widely deployed in public infrastructure and locations. The identification of anomalous incidents such as traffic accidents, robberies or illegal activities is a vital role of video surveillance. Most existing monitoring systems, however, also need human operators and manual inspection (prone to disturbances and tiredness). Therefore, smart computer vision algorithms for automated video anomaly / violence detection are increasingly needed today. A small step towards resolving detection of anomalies is to build algorithms to detect a particular anomalous occurrence, such as violence detector [1], fight action detection [2, 3], and traffic accident detector [4, 5]. Video action recognition has gained increased attention in recent years when achieving very promising performance by taking advantage of CNN’s incredible robustness. © The Author(s) 2020. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as lo
Data Loading...
