TALK DESCRIPTION: Recent advances in the democratization of AI have been enabling the widespread use of generative models, causing the exponential rise of fake content. Nudification of over 680.000 women by a social bot, impersonation scams worth millions of dollars, or spreading political misinformation through synthetic politicians are just the footfall of the deep fake dystopia. As every technology is simultaneously built with its counterpart to neutralize it, this is the perfect time to fortify our eyes with deep fake detectors. Deep fakes depend on photorealism to disable our natural detectors: we cannot simply look at a video to decide if it is real. On the other hand, this realism is not preserved in physiological, biological, and physical signals of deep fakes, yet. In this talk, I will begin with presenting our renowned FakeCatcher, which detects synthetic content in portrait videos using heart beats, as a preventive solution for the emerging threat of deep fakes. Detectors blindly utilizing deep learning are not as effective in catching fake content, as generative models keep producing formidably realistic results. My key assertion follows that such signals hidden in portrait videos can be used as an implicit descriptor of authenticity, like a generalizable watermark of humans, because they are neither spatially nor temporally preserved in deep fakes. Building robust and accurate deep detectors by exhaustively analyzing heartbeats, PPG signals, eye vergence, and gaze movements of deep fake actors reinforce our perception of reality. Moreover, we also innovate novel models to detect the source generator of any deep fake by exploiting its heart beats to unveil residuals of different generative models. Achieving leading results over both existing datasets and our recently introduced in-the-wild dataset justifies our approaches and pioneers a new dimension in deep fake research. SPEAKER BIO: (ACM Distinguished Speaker) Dr. Ilke Demir is based in Hermosa Beach, CA. In the overlap of computer vision and machine learning, Dr. Ilke Demir's research focuses on generative models for digitizing the real world, deep fake detection and generation techniques, analysis and synthesis approaches in geospatial machine learning, and computational geometry for synthesis and fabrication. Currently, she is a Senior Staff Research Scientist at Intel Corporation. Dr. Demir earned her Ph.D. and M.S. in Computer Science from Purdue University advised by Prof. Daniel Aliaga, and her B.S. in Computer Engineering from Middle East Technical University with a minor in Electrical Engineering. Her Ph.D. dissertation conceives geometric and topological shape processing approaches for reconstruction, modeling, and synthesis; which pioneered the area of proceduralization. Afterwards, Dr. Demir joined Facebook as a Postdoctoral Research Scientist working with Prof. Ramesh Raskar from MIT, where their team developed the breakthrough innovation on generative street addresses. Her research further included deep learning approaches for human behavior understanding in next generation virtual reality headsets, geospatial machine learning for map creation, and 3D reconstruction at scale. At the intersection of art and science, Dr. Demir contributed to several animated feature and VR/AR short films in Pixar Animation Studios and Intel Studios, respectively. She established the research foundations of the worldÕs largest volumetric capture studio at Intel, bridging the gap between the creative process and AI approaches. Dr. Demir has been actively involved in women in science organisms, always being an advocate for women and underrepresented minorities. speakers.acm.org/speakers/demir_14351 ilkedemir.github.io For more details: www.meetup.com/SF-Bay-ACM/events/279138953/ 0:00 Chapter Intro 3:11 Presentation Intro 4:53 Presentation 5:49 Deep Generative Model 7:48 Democratization of AI 8:18 Deep Fakes 9:28 Detection of Deep Fakes 11:13 FakeCatcher 25:53 Deep Fake Source Detection 35:33 Deep Fake Detection via Gaze Tracking 42:15 Responsible Deep Fakes: Multi-source Generation 49:44 Responsible Deep Fakes: 3D Digital Humans 51:16 Key Takeaways 52:46 Q&A

Be the first to comment