Built independently by an author, for readers. Read the story and support ChapterPal

keyword

forensic similarity

Forensic similarity is a digital forensics concept and technique that assesses whether two separate image regions share the same intrinsic forensic traces. These traces consist of visually imperceptible statistical patterns and device fingerprints introduced during capture and processing, including sensor noise, color filter array interpolation, and compression artifacts. Instead of classifying whether an image patch matches a specific, predefined editing method or camera model, forensic similarity measures the consistency of the underlying traces between pairs of patches. By identifying localized discrepancies across different areas of an image, this approach enables the detection and localization of tampering, such as image splicing or compositing, even when dealing with previously unseen processing operations or camera sources.

1 item

Robust Image Forgery Detection over Online Social Network Shared Images

Robust Image Forgery Detection over Online Social Network Shared Images

Haiwei Wu, Jiantao Zhou, Jinyu Tian, Jun Liu

OrganizationsUniversity of Macau

Why you should read this

Proposes a training framework that decouples social network transmission artifacts into predictable and adversarial unseen noise, enabling forensic detectors to reliably identify manipulated images degraded by online compression and resizing.

The increasing abuse of image editing softwares, such as Photoshop and Meitu, causes the authenticity of digital images questionable. Meanwhile, the widespread availability of online social networks (OSNs) makes them the dominant channels for transmitting forged images to report fake news, propagate rumors, etc. Unfortunately, various lossy operations adopted by OSNs, e.g., compression and resizing, impose great challenges for implementing the robust image forgery detection. To fight against the OSN-shared forgeries, in this work, a novel robust training scheme is proposed. We first conduct a thorough analysis of the noise introduced by OSNs, and decouple it into two parts, i.e., predictable noise and unseen noise, which are modelled separately. The former simulates the noise introduced by the disclosed (known) operations of OSNs, while the latter is designed to not only complete the previous one, but also take into account the defects of the detector itself. We then incorporate the modelled noise into a robust training framework, significantly improving the robustness of the image forgery detector. Extensive experimental results are presented to validate the superiority of the proposed scheme compared with several state-of-the-art competitors. Finally, to promote the future development of the image forgery detection, we build a public forgeries dataset based on four existing datasets and three most popular OSNs. The designed detector recently won the top ranking in a certificate forgery detection competition^1. The source code and dataset are available at https://github.com/HighwayWu/ImageForensicsOSN.

Added

2026-09-26