Abstract
Moving away from hand-crafted feature extraction, the use of data-driven convolution neural network (CNN)-based algorithms facilitates the realization of end-to-end automated forgery detection in multimedia forensics. On the basis of fingerprints acquired by images from different camera models, the goal of this paper is to design an effective detector capable of completing image forgery detection and localization. Specifically, relying on the designed constant high-pass filter, we first establish a well-performing CNN architecture to adaptively and automatically extract characteristics, and design a reliability fusion map (RFM) to improve localization resolution, and tamper detection accuracy. The extensive results from our empirical experiments demonstrate the effectiveness of our proposed RFM-based detector, and its better performance than other competing approaches.
Highlights
As digital and other communications technologies advance, digital images, videos and audio files can be conveniently acquired from various devices, ranging from the conventional closed-circuit television cameras (CCTVs), digital cameras to other Internet of Things (IoT) devices with image, video and audio capturing capabilities (e.g., Ring Doorbell Camera)
In order to comprehensively evaluate the performance of our proposed reliability fusion map (RFM)-based detector, we focus on pre-processing effectiveness, binary tampering detection, and forgery localization
The resolution of forgery localization is becoming more challenging for digital image forensics
Summary
As digital and other communications technologies advance, digital images, videos and audio files can be conveniently acquired from various devices, ranging from the conventional closed-circuit television cameras (CCTVs), digital cameras to other Internet of Things (IoT) devices with image, video and audio capturing capabilities (e.g., Ring Doorbell Camera). Modifying an image has become easier, due to the availability of inexpensive image, video and audio (collectively referred to as multimedia) editing software. Implications of forged multimedia files, for example using re-sampling [1,2] or copy-moving [3,4], include ownership infringement or fraudulent activities. $243,000) to the bank account of a Hungarian supplier” (https://www.forbes.com/sites/jessedamiani/2019/09/03/a-voice-deepfakewas-used-to-scam-a-ceo-out-of-243000/). This necessitates the need to design an effective and robust forensic detector with the capability of providing reliable digital evidence. The study of both source identification and tampering detection is a relatively mature topic [5,6,7]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
