Capturing the World Through Hidden Signals in Video
Seminar | April 4 | 1-2 p.m. | Soda Hall, 430-438 Wozniak Lounge
Abe Davis, Postdoctoral Researcher, Stanford University
The world is full of visual signals that go largely unnoticed; some are too subtle or fast for us to see, while others simply occur in patterns we dont recognize. I will show how insights from physics, art, and engineering can help us identify such signals, recover them from video, and use them to model real-world problems. The first part of my talk will focus on visual vibration analysis, where I will show how perceptually invisible motion can be used to recover sound from silent video, monitor the health of aging infrastructure, and build interactive physical simulations of visible objects. Then I will show how, by leveraging theory and expertise from various arts (e.g., photography, cinematography, music, and dance) we can use hidden visual signals to build powerful and automated interactive tools for artists.
Abe Davis is a postdoctoral researcher at Stanford University working at the intersections of computer graphics, vision, HCI, and civil engineering. He earned his Ph.D. in Electrical Engineering and Computer Science from MIT in 2016 and is the recipient of the MIT Sprowls Award for Outstanding Dissertation in Computer Science and the ACM SIGGRAPH Outstanding Doctoral Dissertation Honorable Mention Award. Abe was awarded NSF and Mathworks graduate fellowships, named one of Forbes Magazine's "30 under 30", Business Insider's "50 Scientists Who are Changing the World" and "8 Innovative Scientists in Tech and Engineering", and won the "Most Practical SHM Solution for Civil Infrastructures" Award at IWSHM 2017. He is currently funded by a Magic Grant from the Brown Institute for Media Innovation.