Faculty Candidate & CSE 600 talk with James Tompkin from Harvard University.

Title: Intelligent Video Interfaces

Video best represents how we visually perceive, and with ubiquitous cameras everyone can both capture their world and be captured by their environment. However, our tools to interact with video often require laborious work or expert skill, which limits our ability for natural human-computer interaction. With advances in image understanding through computer graphics and vision, we can create intelligent tools for video interaction: First, to transform video into a creative and connected medium, so that non-experts can make sophisticated content-based edits and explore large collections in context-aware ways. Second, to transform video cameras from passive observers into essential interaction tools, so that we can analyze human motion for real-time creation and control of digital imagery.

Bio:
James Tompkin is a post-doctoral fellow in the Harvard University visual computing group of Hanspeter Pfister. He received his EngD in virtual environments, imaging, and visualization from University College London under the supervision of Jan Kautz, and then was a post-doctoral researcher at Max-Planck-Institute for Informatics with Christian Theobalt. He has been sponsored by BBC Research and Intel IVCI, and has interned at Disney Research with Wojciech Matusik. Through removing barriers from the creative process with new interactive tools, his collaborations with architects and artists have been exhibited at venues such as The Building Center in London and the Museum of the Moving Image in New York City.