Real-Time Hand Tracking Using a Sum of Anisotropic Gaussians Model
Srinath Sridhar, Helge Rhodin, Hans‐Peter Seidel, Antti Oulasvirta, Christian Theobalt · MPG.PuRe (Max Planck Society) · 2014
Real-time marker-less hand tracking is of increasing im-portance in human-computer interaction. Robust and ac-curate tracking of arbitrary hand motion is a challenging problem due to the many degrees of freedom, frequent self-occlusions, fast motions, and uniform skin color. In this pa-per, we propose a new approach that tracks the full skeleton motion of the hand from multiple RGB cameras in real-time. The main contributions include a new generative tracking method which employs an implicit hand shape representa-tion based on Sum of Anisotropic Gaussians (SAG), and a pose fitting energy that is smooth and analytically differen-tiable making fast gradient based pose optimization possi-ble. This shape representation, together with a full perspec-tive projection model, enables more accurate hand mod-eling than a related baseline method from literature. Our method achieves better accuracy than previous methods and runs at 25 fps. We show these improvements both qualita-tively and quantitatively on publicly available datasets. 1.