报告人：David Doermann Professor，IEEE and IAPR fellow
Dr. David Doermann is a Professor of Empire Innovation at the University at Buffalo (UB). David is a leading research and innovative thinker in the areas of document image analysis and recognition. He and his group of researchers focused on many innovative topics related to analysis and processing of document images and video including triage, visual indexing and retrieval, enhancement and recognition of both textual and structural components of visual media. David has over 250 publications in conferences and journals, is a fellow of the IEEE and IAPR,has numerous awards including an honorary doctorate from the University of Oulu,Finland and is a founding Editor-in-Chief of the International Journal on DocumentAnalysis and Recognition.
Tampering with or misrepresenting content in images and video is a concern for many groups. The ability to create or manipulate this content, which we typically trust because we can “see” it, is becoming more pervasive. Until recently, only high value media has been analyzed for integrity, often done by hand and with adhoc tools. While the research community has shown some progress in developing automated tools that indicate tampering, localize potential manipulations and document inconsistencies, they almost universally require human judgement for decision making. This talk will first report on progress and lay out the successes and technical achievements.The second part of this talk will focus on the role of adversarial networks for both hardening our algorithms, and the threat caused by their ability to automatically generate and modify content. He will overview these networks and give examples of how they are used in Media Forensics.Finally, he will provide some summary thoughts on how other Artificial Intelligence technologies are being applied and the challenges of trust, ethical behavior and security.