An On-Line and Adaptive Method for Detecting Abnormal Events in Videos Using Spatio-Temporal ConvNet

被引：26

作者：

Bouindour, Samir ^{[1
]}

Snoussi, Hichem ^{[1
]}

Hittawe, Mohamad Mazen ^{[2
]}

Tazi, Nacef ^{[1
]}

Wang, Tian ^{[3
]}

机构：

[1] Univ Technol Troyes, CNRS, ICD LM2S, F-10000 Troyes, France

[2] King Abdullah Univ Sci & Technol, CEMSE, Thuwal 239556900, Saudi Arabia

[3] Beihang Univ, Sch Automat Sci & Elect Engn, Beijing 100083, Peoples R China

来源：

APPLIED SCIENCES-BASEL | 2019年 / 9卷 / 04期

关键词：

abnormal event detection; deep learning; Convolutional Neural Network; unsupervised learning; online; adaptive systems; ANOMALY DETECTION; NEURAL-NETWORKS; LOCALIZATION; BEHAVIOR; IMAGES;

D O I：

10.3390/app9040757

中图分类号：

O6 [化学];

学科分类号：

0703 ;

摘要：

We address in this paper the problem of abnormal event detection in video-surveillance. In this context, we use only normal events as training samples. We propose to use a modified version of pretrained 3D residual convolutional network to extract spatio-temporal features, and we develop a robust classifier based on the selection of vectors of interest. It is able to learn the normal behavior model and detect potentially dangerous abnormal events. This unsupervised method prevents the marginalization of normal events that occur rarely during the training phase since it minimizes redundancy information, and adapt to the appearance of new normal events that occur during the testing phase. Experimental results on challenging datasets show the superiority of the proposed method compared to the state of the art in both frame-level and pixel-level in anomaly detection task.

引用

页数：17