GANerated Hands for Real-Time 3D Hand Tracking from Monocular RGB —Extended 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 enhancing the training data for our CNN we propose 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 sogenerated images follow the same statistical distribution as real-world hand images while preserving geometric properties (such as hand pose). We demonstrate that our hand tracking system outperforms the current state-of-the-art on challenging RGB-only footage.