GANerated Hands for Real-time 3D Hand Tracking from Monocular RGB

• Published in: arXiv.org
• Main Authors: Mueller, Franziska, Bernard, Florian, Sotnychenko, Oleksandr, Mehta, Dushyant, Sridhar, Srinath, Casas, Dan, Theobalt, Christian
• Format: Paper
• Language: English
• Published: Ithaca Cornell University Library, arXiv.org 04.12.2017
• Subjects: Artificial neural networks; Consistency; Hand (anatomy); Neural networks; Real time; Three dimensional models; Tracking systems; Training

We address the highly challenging problem of real-time 3D hand tracking based on a monocular RGB-only sequence. Our tracking method combines a convolutional neural network with a kinematic 3D hand model, such that it generalizes well to unseen data, is robust to occlusions and varying camera viewpoi...