6211 Sudikoff Laboratory
Hanover, NH 03755, USA
Email: wei DOT fan AT dartmouth DOT edu
I've been working on both image anti-forensics and forensics. I also used to work on reversible image watermarking, which is still within the domain of multimedia security at large. In the following, I will give a very brief introduction about image forensics and anti-forensics. If you are interested in my work, please check my Publications page for details. There, you can find the full papers (or their links), Matlab source codes, and supplementary materials.
Seeing is believing, or a picture is worth a thousand words, they always say. That is probably why the digital image is one of the most commonly used multimedia types on Internet. According to Mary Meeker’s annual Internet trends report , over 1.8 billion photos are uploaded and shared on Internet per day! Digital images are literally ubiquitous, in which we historically have trust. However, this confidence is being constantly shaken due to the widespread availability of high-quality cameras and powerful photo-editing tools. In the figure below, the Girl with A Pearl Earring, portrayed by the 17-century painter Johannes Vermeer, appears to appreciate the music coming from the beats.1↩ Can you believe your eyes?
(a) Girl with a pearl earring
(b) Girl with the beats
Figure: (a) is the Girl with A Pearl Earring painted by Johannes Vermeer around 1665. The beats, a product of modern technologies, were integrated into the digital version of this classic painting by the Worth1000 website user bigchopper in (b).
The doubt thrown upon digital images has urged the development of digital image forensics, trying to restore some trust to digital images. The main objectives of forensics are to analyze a given digital image so as to detect whether it is a forgery, to identify its origin, to trace its processing history, or to reveal latent details invisible to human naked eyes .
Similar to cryptography vs. cryptanalysis and steganography vs. steganalysis, the two sides of the coin regarding image authentication are forensics and anti-forensics. Digital image anti-forensics is the technique to expose the limitations of forensic methods, with the ultimate goal to develop more trustworthy forensics . In literature, anti-forensics is also synonymously referred to as counter-forensics. This research direction is important, since we have to design against possible anti-forensics for reliable forensics. The main objective of image anti-forensics is to perform certain operations to digital images, disguising traces left by image editing so that they will no longer be detected by forensic algorithms.