Object-based forgery detection in surveillance video using capsule network

Abstract

Due to the rapid growth of digital multimedia technology, the importance of multimedia data such as digital images/videos is increasing rapidly in diverse domains. This includes surveillance CCTV footages, which act as primary evidences in many court cases, including highly sensitive contexts. Today, with the wide availability of low-cost image/video manipulating software, digital images/videos have become highly vulnerable to manipulation/modification attacks; one such attack is the class of object-based forgery in surveillance videos. In this paper, we propose a Capsule Network based digital forensic technique for detection of object-based forgery in surveillance videos. In the proposed technique, we use motion residual, computed from every video frame, to extract intra-and inter-frame inherent statistical characteristics of the video sequence, as the input of capsule network. Our experimental results indicate that the proposed technique achieves significant performance in terms of authentic, double compressed and forged frame detection, irrespective of the group of pictures length and degree of compression in videos.

Real‐time object‐removal tampering localization in surveillance videos by employing YOLO‐V8

“…area can be seen in Figure 2. The obtained results are better than those reported in [16,17]. With this strategy, we can obtain information about moving objects and static background information.…”

“…The computed motion residue provides better forged area localization inside a modified frame. The suggested algorithms perform better than [4,16,17,31]. We can see visual representations of the suggested method in Figure 2.…”

“…This algorithm does not give the precise localization of the tampered region. Bakas et al [16] suggested a capsule network for object-based forgery in surveillance videos on processed frames.…”

“…The comparative study of the proposed method can be seen in Table 5. The overall accuracy (mAP = 0.99) of the system is better than that of [16,17,28], while the precision is lagging by a value of 0.01 concerning [11,24]. The metrics behavior in Table 2 represents the proper training of the proposed system, as losses are decreasing and accuracy is increasing.…”

“…Bakas et al. [16] suggested a capsule network for object‐based forgery in surveillance videos on processed frames. The proposed algorithm in this method computes the motion residue using a collision operator, which gives a rough localization of the forged region.…”