Combinatorical and Algebraic Structures Seminar

Session details

Date: 6.12.2016
Speaker: Tomáš Denemark, ECE, SUNY Binghamton
Title: Image Steganography and Steganalysis with Side-Information
Abstract: Steganography studies the art of hidden communication. By carefully introducing small changes to pixel colors, images can be modified to carry secret messages. The most important attribute of steganography is statistical undetectability -- the hidden message must be hidden so that its very presence is difficult (impossible) to establish. Methods of statistical hypothesis testing, signal detection, and estimation coupled with artificial intelligence are employed to detect steganographic communication. In my talk, I explain how various sources of side-information at the sender and at the detector can be used to increase steganographic security as well as improve its detection.

Return to index.