软件导刊2024，Vol.23Issue(10)：187-192,6.DOI:10.11907/rjdk.232013
基于深度可分离卷积的异常驱动视频异常检测
Anomaly-Driven Video Anomaly Detection Based on Depth-Separable Convolution
摘要
Abstract
Video anomaly detection has become a hot issue in current research with profound practical application value.Aiming at the prob-lems of high computational complexity of 3D convolution in video anomaly detection,difficulty in training,and easy overfitting by utilizing on-ly normal data when using reconstruction methods for detection,a novel deeply deparable convolutional anomaly-driven network is proposed.The network firstly extracts jump frames as pseudo-anomaly samples through manual feature extraction to assist training,secondly designs the deeply deparable convolutional network to reduce the number of computational parameters for 3D convolution,and finally allows the network to learn to differentiate between anomalous and normal data by minimizing the reconstruction error of normal data and maximizing the anoma-lous data.Experimental results show that the model exhibits competitive performance on all major public datasets,with accuracy rates of 91.3%,99.2%,87.4%and 98.6%on UCSDped1,UCSDped2,Avenue and UMN datasets,respectively.In addition,the model has strong sensitivity to anomaly detection,and has strong generalization ability and robustness.关键词
深度可分离卷积/伪异常/帧重构/视频异常检测Key words
deeply separable convolution/pseudo-anomaly/frame reconstruction/video anomalies detection分类
信息技术与安全科学引用本文复制引用
李新,宋刘广,孙钰奇,曾佳全..基于深度可分离卷积的异常驱动视频异常检测[J].软件导刊,2024,23(10):187-192,6.基金项目
广西科技计划项目(2020GXNSFAA297255) （2020GXNSFAA297255）
广西嵌入式技术与智能系统重点实验室开放基金项目(2020-2-6) （2020-2-6）
