Cognition Guided Video Anomaly Detection Framework for Surveillance Services

被引:0
作者
Zhang, Menghao [1 ]
Wang, Jingyu [1 ]
Qi, Qi [1 ]
Zhuang, Zirui [1 ]
Sun, Haifeng [1 ]
Liao, Jianxin [1 ]
机构
[1] Beijing Univ Posts & Telecommun, State Key Lab Networking & Switching Technol, Beijing 100876, Peoples R China
基金
中国国家自然科学基金;
关键词
Anomaly detection; Surveillance; Knowledge engineering; Task analysis; Visualization; Explosions; Data models; Multi-layer GCN; prior knowledge; prompt tuning; video anomaly detection; surveillance service;
D O I
10.1109/TSC.2024.3407588
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The aim of surveillance services is to detect anomalous events that occur in given surveillance videos. Most existing video anomaly detection methods rely on minimizing reconstruction or prediction errors due to the lack of abnormal data, which results in poor generalization and overfitting. In fact, cognitions for anomalies in surveillance videos mainly relies on crucial relationships, including ones between objects and ones between objects and scenes. Focusing on this property of anomaly detection, a Cognition Guided Video Anomaly Detection framework based on prior knowledge is proposed, called CG-VAD. CG-VAD introduces both explicit and implicit prior knowledge into the frame prediction network to let the model exploit crucial relationships. Explicit knowledge containing crucial relationships related to anomaly is introduced into the anomaly detection model through a proposed embedding network based on multi-layer Graph Convolutional Networks. Implicit knowledge in the form of learnable parameters enhances the ability of the model to learn crucial relationships through prompt tuning. By integrating prior knowledge to focus the model on the relationships associated with the anomaly, we find that CG-VAD is not only quick to adapt to new real-world scenarios, but it is also able to recognize the type of anomaly. We have conducted extensive experiments on four benchmark datasets and the results indicate that the proposed method outperforms previous methods. Specifically, CG-VAD achieves an AUROC score of 87.2$\%$% on the ShanghaiTech dataset.
引用
收藏
页码:2109 / 2123
页数:15
相关论文
共 81 条
[1]   Exploring Long Tail Visual Relationship Recognition with Large Vocabulary [J].
Abdelkarim, Sherif ;
Agarwal, Aniket ;
Achlioptas, Panos ;
Chen, Jun ;
Huang, Jiaji ;
Li, Boyang ;
Church, Kenneth ;
Elhoseiny, Mohamed .
2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021), 2021, :15901-15910
[2]   UBnormal: New Benchmark for Supervised Open-Set Video Anomaly Detection [J].
Acsintoae, Andra ;
Florescu, Andrei ;
Georgescu, Mariana-Iuliana ;
Mare, Tudor ;
Sumedrea, Paul ;
Ionescu, Radu Tudor ;
Khan, Fahad Shahbaz ;
Shah, Mubarak .
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022), 2022, :20111-20121
[3]   Cross-Domain Video Anomaly Detection without Target Domain Adaptation [J].
Aich, Abhishek ;
Peng, Kuan-Chuan ;
Roy-Chowdhury, Amit K. .
2023 IEEE/CVF WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV), 2023, :2578-2590
[4]   Hierarchical Scene Normality-Binding Modeling for Anomaly Detection in Surveillance Videos [J].
Bao, Qianyue ;
Liu, Fang ;
Liu, Yang ;
Jiao, Licheng ;
Liu, Xu ;
Li, Lingling .
PROCEEDINGS OF THE 30TH ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA, MM 2022, 2022, :6103-6112
[5]   SSMTL plus plus : Revisiting self-supervised multi-task learning for video anomaly detection [J].
Barbalau, Antonio ;
Ionescu, Radu Tudor ;
Georgescu, Mariana-Iuliana ;
Dueholm, Jacob ;
Ramachandra, Bharathkumar ;
Nasrollahi, Kamal ;
Khan, Fahad Shahbaz ;
Moeslund, Thomas B. ;
Shah, Mubarak .
COMPUTER VISION AND IMAGE UNDERSTANDING, 2023, 229
[6]  
Cao Hu, 2023, Computer Vision - ECCV 2022 Workshops: Proceedings. Lecture Notes in Computer Science (13803), P205, DOI 10.1007/978-3-031-25066-8_9
[7]   Video anomaly detection with spatio-temporal dissociation [J].
Chang, Yunpeng ;
Tu, Zhigang ;
Xie, Wei ;
Luo, Bin ;
Zhang, Shifu ;
Sui, Haigang ;
Yuan, Junsong .
PATTERN RECOGNITION, 2022, 122
[8]  
Chen CW, 2022, AAAI CONF ARTIF INTE, P230
[9]   Multi-Scale LSTM Model for BGP Anomaly Classification [J].
Cheng, Min ;
Li, Qing ;
Lv, Jianming ;
Liu, Wenyin ;
Wang, Jianping .
IEEE TRANSACTIONS ON SERVICES COMPUTING, 2021, 14 (03) :765-778
[10]   InfoGCN: Representation Learning for Human Skeleton-based Action Recognition [J].
Chi, Hyung-gun ;
Ha, Myoung Hoon ;
Chi, Seunggeun ;
Lee, Sang Wan ;
Huang, Qixing ;
Ramani, Karthik .
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022), 2022, :20154-20164