Reverse erasure guided spatio-temporal autoencoder with compact feature representation for video anomaly detection  

作　　者：Yuanhong ZHONG Xia CHEN Jinyang JIANG Fan REN 

机构地区：[1]School of Microelectronics and Communication Engineering,Chongqing University,Chongqing,400044,China [2]State Grid Chongqing Electric Power Research Institute,Chongqing,401123,China [3]Changan Software Technology Company,Changan Automobile Corp.,Chongqing,401120,China

出　　处：《Science China(Information Sciences)》2022年第9期282-284,共3页中国科学（信息科学）（英文版）

基　　金：partially supported by Special Project of Technological Innovation and Application Development of Chongqing(Grant No.cstc2019jscx-msxmX0167);National Natural Science Foundation of China(Grant No.61501069)。

摘　　要：Video anomaly detection aims to learn normal patterns and identify the samples deviating from normal patterns as anomalies.In early research,methods based on handcrafted low-level features have been widely studied.However,the representation power of low-level features is insufficient for describing various patterns,causing a bottleneck in handcrafted-feature-based anomaly detection.

关 键 词：representation detection CRAFT 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP18[自动化与计算机技术—计算机科学与技术]