- This event has passed.
Media Forensics and Ethical Artificial Intelligence
November 17, 2021 @ 2:30 pm - 3:30 pm
Misinformation and disinformation have existed ever since information itself existed, whether it was a state-sponsored actor trying to sow political/civil unrest, manipulate elections or undermine democracies, or simply unintentional ignorance and tribalism. However, rapid advances in artificial intelligence, especially generative models such as generative adversarial networks, make it very easy to produce multimedia artifacts (images, video and audio) that show people doing and saying what they never did or said. The widespread use of manipulated and generated multimedia as tools of misinformation on social networking platforms has major societal consequences, including child safety, public opinion and democracy and criminal justice. The prevalence of fabricated multimedia necessitates a new wave of forensic computer vision methods for detecting and localizing image and video manipulations. In this talk, I will give a brief overview of current challenges in multimedia forensics; then, I will present two novel multimedia forensics methods developed by my group. First, I will present ManTraNet (Manipulation Tracing Network), a unified deep convolutional neural network architecture for detecting and localizing various types of image manipulations, such as copy-move and splicing. Second, I will present a state-of-the-art deep learning pipeline for detecting deepfake videos. Our pipeline demonstrates state-of-the-art evaluation results on several standard benchmarks. Finally, I will wrap up with conclusions and thoughts for future directions in multimedia forensics research and ethical artificial intelligence in general. Co-sponsored by: University of Nevada, Reno Speaker(s): Dr. Wael Abd-Almageed, 1664 N Virginia St, Reno, Nevada, United States, 89557, Virtual: https://events.vtools.ieee.org/m/289817