Abstract

Analyzing human motion is a challenging task with a wide variety of applications in computer vision and in graphics. One such application, of particular importance in computer animation, is the retargeting of motion from one performer to another. While humans move in three dimensions, the vast majority of human motions are captured using video, requiring 2D-to-3D pose and camera recovery, before existing retargeting approaches may be applied. In this talk, we will present a new method for retargeting video-captured motion between different human performers, without the need to explicitly reconstruct 3D poses and/or camera parameters. Our key idea is to train a deep neural network to decompose temporal sequences of 2D poses into three components: motion, skeleton, and camera view-angle. In this way, we learn to extract, directly from a video, a high-level latent motion representation, which is invariant to the skeleton geometry and the camera view. Having extracted such a representation, we are able to re-combine motion with novel skeletons and camera views, and decode a retargeted temporal sequence, which we compare to a ground truth from a synthetic dataset. We demonstrate that our framework can be used to robustly extract human motion from videos, bypassing 3D reconstruction, and outperforming existing retargeting methods, when applied to videos in-the-wild. It also enables additional applications, such as performance cloning, video-driven cartoons, and motion retrieval. 

Bio

Kfir is a researcher in the Reality Capture Group at the Advanced InnovationCenter for Future Visual Entertainment (AICFVE) in Beijing. In parallel, Kfir is pursuing his Ph.D in Tel-Aviv University, under the supervision of Prof. Daniel Cohen-Or, where his areas of interests include deep neural networks architectures for computer graphics application. His B.Sc. (summa cum laude) and M.Sc. (cum laude) in Electrical Engineering from the Technion were done under the supervision of Prof. Yonina C. Eldar in the area of signal processing.