GANerated Hands for Real-Time 3D Hand Tracking from Monocular RGBF. MuellerF. BernardO. SotnychenkoD. MehtaS. SridharD. CasasC. TheobaltSpotlight @ Computer Vision and Pattern Recognition (CVPR) 2018
We propose a novel method for real-time hand tracking from monocular RGB input which combines a CNN-based regressor and a kinematic optimization framework. To enhance our synthetic training data, we introduce a new geometrically consistent image-to-image translator for unpaired examples.
Real-time Hand Tracking under Occlusion from an Egocentric RGB-D SensorF. MuellerD. MehtaO. SotnychenkoS. SridharD. CasasC. TheobaltInternational Conference on Computer Vision (ICCV) 2017
We present a method for real-time hand tracking under occlusion in cluttered egocentric scenes from a single RGB-D camera. To enable training of our machine learning components, we introduce a new large-scale dataset SynthHands which was captured using a mixed reality approach. Furthermore, we propose a real benchmark dataset EgoDexter which provides annotated fingertip positions.
April 2015 - March 2016:
Master Studies in Computer Science at Saarland University, Saarbrücken, Germany
Title of Master's Thesis: Real-Time Hand-Object Tracking Using a Single Depth Camera (supervisor: Prof. Dr. Christian Theobalt) [pdf]
October 2012 - March 2015:
Bachelor Studies in Computer Science at Saarland University, Saarbrücken, Germany
Title of Bachelor's Thesis: Real-Time Hand Tracking Using Hybrid Pose Optimization (supervisor: Prof. Dr. Christian Theobalt) [pdf]
Abitur at the Warndtgymnasium Völklingen, Germany