Identification of Anomaly Scenes in Videos Using Graph Neural Networks

作     者：Khalid Masood Mahmoud M.Al-Sakhnini Waqas Nawaz Tauqeer Faiz Abdul Salam Mohammad Hamza Kashif 

作者机构：Department of Software EngineeringLahore Garrison UniversityLahore54000Pakistan School of BusinessSkyline University CollegeSharjahUAE Al-Madinah International University–Faculty of Computer and Information TechnologyKuala LumpurMalaysia College of Computer and Information SystemsIslamic University of MadinahMadinahKingdom of Saudi Arabia School of ITSkyline University CollegeSharjahUAE Faculty of EngineeringComputing and Science(FECS)-Swinburne University of TechnologySarawak Department of General EducationSkyline University CollegeUniversity City of SharjahUAE Department of Computer ScienceLahore Garrison UniversityLahore54000Pakistan 

出 版 物：《Computers, Materials & Continua》 (计算机、材料和连续体（英文）)

年 卷 期：2023年第74卷第3期

页      面：5417-5430页

核心收录：

学科分类：08[工学] 080203[工学-机械设计及理论] 0802[工学-机械工程] 

主　　题：Graph neural network deep learning anomaly detection auto encoders 

摘      要：Generally,conventional methods for anomaly detection rely on clustering,proximity,or *** themassive growth in surveillance videos,outliers or anomalies find ingenious ways to obscure themselves in the network and make conventional techniques *** research explores the structure of Graph neural networks(GNNs)that generalize deep learning frameworks to graph-structured *** node in the graph structure is labeled and anomalies,represented by unlabeled nodes,are predicted by performing random walks on the node-based graph *** to their strong learning abilities,GNNs gained popularity in various domains such as natural language processing,social network analytics and *** detection is a challenging task in computer vision but the proposed algorithm using GNNs efficiently performs the identification of *** Graph-based deep learning networks are designed to predict unknown objects and *** our case,they detect unusual objects in the form of malicious *** edges between nodes represent a relationship of nodes among each *** case of anomaly,such as the bike rider in Pedestrians data,the rider node has a negative value for the edge and it is identified as an *** encoding and decoding layers are crucial for determining how statistical measurements affect anomaly identification and for correcting the graph path to the best possible *** show that the proposed framework is a step ahead of the traditional approaches in detecting unusual activities,which shows a huge potential in automatically monitoring surveillance *** autonomous monitoring of CCTV,crime control and damage or destruction by a group of people or crowd can be identified and alarms may be triggered in unusual activities in streets or public *** suggested GNN model improves accuracy by 4%for the Pedestrian 2 dataset and 12%for the Pedestrian 1 dataset compared to a few state-of the-art techniq