2019/06/15 by Pan Jonathan, Pan, Jonathan
Computer Science · #Anomaly Detection Techniques and Applications #Digital Media Forensic Detection #Network Security and Intrusion Detection
paper · pdf · doi:10.48550/arxiv.1906.06475
Surveillance cameras, which is a form of Cyber Physical System, are deployed\nextensively to provide visual surveillance monitoring of activities of interest\nor anomalies. However, these cameras are at risks of physical security attacks\nagainst their physical attributes or configuration like tampering of their\nrecording coverage, camera positions or recording configurations like focus and\nzoom factors. Such adversarial alteration of physical configuration could also\nbe invoked through cyber security attacks against the camera's software\nvulnerabilities to administratively change the camera's physical configuration\nsettings. When such Cyber Physical attacks occur, they affect the integrity of\nthe targeted cameras that would in turn render these cameras ineffective in\nfulfilling the intended security functions. There is a significant measure of\nresearch work in detection mechanisms of cyber-attacks against these Cyber\nPhysical devices, however it is understudied area with such mechanisms against\nintegrity attacks on physical configuration. This research proposes the use of\nthe novel use of deep learning algorithms to detect such physical attacks\noriginating from cyber or physical spaces. Additionally, we proposed the novel\nuse of deep learning-based video frame interpolation for such detection that\nhas comparatively better performance to other anomaly detectors in\nspatiotemporal environments.\n