Vision & Graphics Lab (Digitization & Computer Vision and Graphics)

The Academy Award-winning Vision & Graphics Lab has been bringing deep learning research and novel data-driven techniques to advance end-to-end acquisition and content creation, with new capabilities such as deployablilty in the field, full body digitization, and differentiable renderings, and true-to-life real-time performance capture. 

VGL has been developing fundamental research for emerging applications such as real-time teleportation, interactions with virtual avatars, and VR and AR simulations enabling new training, education capabilities as well as novel command-level decision making tools relevant to the Army’s STE efforts. 

The work on synthesizing photorealistic digital humans using deep neural networks is also impacting media forensics research, in particular supporting the detection of deep fakes, as demonstrated in our cutting-edge analysis tools based on soft biometrics.