English
 
User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Paper

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

MPS-Authors
/persons/resource/persons134216

Mueller,  Franziska
Computer Graphics, MPI for Informatics, Max Planck Society;

/persons/resource/persons214986

Bernard,  Florian
Computer Graphics, MPI for Informatics, Max Planck Society;

/persons/resource/persons199773

Sotnychenko,  Oleksandr
Computer Graphics, MPI for Informatics, Max Planck Society;

/persons/resource/persons129023

Mehta,  Dushyant
Computer Graphics, MPI for Informatics, Max Planck Society;

/persons/resource/persons45610

Theobalt,  Christian
Computer Graphics, MPI for Informatics, Max Planck Society;

Locator
There are no locators available
Fulltext (public)

arXiv:1712.01057.pdf
(Preprint), 8MB

Supplementary Material (public)
There is no public supplementary material available
Citation

Mueller, F., Bernard, F., Sotnychenko, O., Mehta, D., Sridhar, S., Casas, D., et al. (2017). GANerated Hands for Real-time 3D Hand Tracking from Monocular RGB. Retrieved from http://arxiv.org/abs/1712.01057.


Cite as: http://hdl.handle.net/21.11116/0000-0000-6156-2
Abstract
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 viewpoints, and leads to anatomically plausible as well as temporally smooth hand motions. For training our CNN we propose a novel approach for the synthetic generation of training data that is based on a geometrically consistent image-to-image translation network. To be more specific, we use a neural network that translates synthetic images to "real" images, such that the so-generated images follow the same statistical distribution as real-world hand images. For training this translation network we combine an adversarial loss and a cycle-consistency loss with a geometric consistency loss in order to preserve geometric properties (such as hand pose) during translation. We demonstrate that our hand tracking system outperforms the current state-of-the-art on challenging RGB-only footage.