Helge Rhodin
Scientist, computer vision and graphics
EPFL IC IINFCOM CVLAB
BC 302 (Bâtiment BC)
Station 14
CH-1015 Lausanne
helge.rhodin@epfl.ch
I am a lecturer at EPFL and postdoctoral researcher in the Computer Vision Lab of Pascal Fua. My research interests range from low-level computer vision, over machine learning, to computer graphics and augmented reality. With the aim of enabling tight, yet rich computer-human interaction, I advance 3D vision algorithms to yield high-quality dynamic reconstructions of our everyday environment from plain video input. For instance,
- Self-supervised training from multi-view images (CVPR 2018, ECCV 2018)
- Real-time human pose estimation from a single camera (SIGGRAPH 2017)
- Egocentric motion capture with applications to VR (SIGGRAPH Asia 2016)
- Fully automatic actor model creation and motion capture (ECCV 2016, TOG 2018)
- A differentiable visibility model for human motion capture from sparse camera setups (ICCV 2015)
- Dynamic mapping of human gestures to virtual characters (EG 2014 and SIGGRAPH Asia 2015)
The latest VR and AR devices offer incredible display capabilities; my 3D vision projects complement them to aid mutual human-computer interaction in our everyday life.
I grew up in Hamburg and received the BSc and MSc degree in Computer science from Saarland University, respectively supervised by Philipp Slusallek and Sebastian Hack, and Michael Breuß and Joachim Weickert.
In 2016 I obtained my PhD at the Max-Planck Institute for Informatics for my work on Motion Capture for Interactive Virtual Worlds, mentored by Christian Theobalt and Hans-Peter Seidel.
Publications
Upcoming
 |
*New* What Face and Body Shapes Can Tell About Height
Semih Günel, Helge Rhodin, Pascal Fua arXiv 2018 [paper pdf] |
 |
*New* Mo2Cap2: Real-time Mobile 3D Motion Capture with a Cap-Mounted Fisheye Camera
Weipeng Xu, Avishek Chatterjee, Michael Zollhöfer, Helge Rhodin, Pascal Fua, Hans-Peter Seidel, Christian Theobalt arXiv 2018 [paper pdf] |
2018
 |
*New* Unsupervised Geometry-Aware Representation Learning for 3D Human Pose Estimation
Helge Rhodin, Mathieu Salzmann and Pascal Fua ECCV 2018 (Oral presentation) [project page] [paper pdf] |
 |
Learning Monocular 3D Human Pose Estimation from Multi-view Images
Helge Rhodin, Jörg Spörri, Isinsu Katircioglu, Victor Constantin,
Frédéric Meyer, Erich Müller, Mathieu Salzmann and Pascal Fua CVPR 2018 (Spotlight presentation) [project page] [paper pdf] [video] |
 |
MonoPerfCap: Human Performance Capture from Monocular Video
Weipeng Xu, Avishek Chatterjee, Michael Zollhöfer, Helge Rhodin, Dushyant Mehta, Hans-Peter Seidel, Christian Theobalt TOG 2018 (Oral presentation at SIGGRAPH 2018) [project page] [paper pdf] [video] |
2017
 |
Monocular 3D Human Pose Estimation In The Wild Using Improved CNN Supervision
Dushyant Mehta, Helge Rhodin, Dan Casas, Pascal Fua, Oleksandr Sotnychenko, Weipeng Xu, and Christian Theobalt 3DV 2017 (Oral presentation) [project page] [dataset] [paper pdf] |
 |
VNect: Real-time 3D Human Pose Estimation with a Single RGB Camera
Dushyant Mehta, Srinath Sridhar, Oleksandr Sotnychenko, Helge Rhodin, Mohammad Shafiei, Hans-Peter Seidel, Weipeng Xu, Dan Casas, and Christian Theobalt
SIGGRAPH 2017 (Oral presentation) [project page] [paper pdf] [video] [CVPR 2017 demo project page] |
2016
 |
From Motion Capture to Interactive Virtual Worlds
Towards Unconstrained Motion-capture Algorithms for Real-time Performance-driven Character Animation
Helge Rhodin PhD Dissertation, Saarland University, 2016 [thesis pdf] |
 |
EgoCap: Egocentric Marker-less Motion Capture with Two Fisheye Cameras
Helge Rhodin, Christian Richardt, Dan Casas, Eldar Insafutdinov, Mohammad Shafiei, Hans-Peter Seidel, Bernt Schiele, and Christian Theobalt SIGGRAPH Asia 2016 (Oral presentation) [project page] [dataset] [paper pdf] [video] |
 |
Model-based Outdoor Performance Capture Nadia Robertini, Dan Casas, Helge Rhodin, Hans-Peter Seidel, and Christian Theobalt 3DV 2016 [project page] [paper pdf] [video] |
 |
General Automatic Human Shape and Motion Capture Using Volumetric Contour Cues Helge Rhodin, Nadia Robertini, Dan Casas, Christian Richardt, Hans-Peter Seidel and Christian Theobalt ECCV 2016 (Spotlight presentation) [project page] [paper pdf] [video] |
2015
 |
A Versatile Scene Model with Differentiable Visibility Applied to Generative Pose Estimation Helge Rhodin, Nadia Robertini, Christian Richardt, Hans-Peter Seidel, and Christian Theobalt ICCV 2015 [project page] [paper pdf] [video] |
 |
Generalizing Wave Gestures from Sparse Examples for Real-time Character Control Helge Rhodin, James Tompkin, Kwang In Kim, Edilson de Aguiar, Hanspeter Pfister, Hans-Peter Seidel, and Christian Theobalt SIGGRAPH Asia 2015 (Oral presentation) [project page] [paper pdf] [video] |
2014
 |
Real-time Hand Tracking Using a Sum of Anisotropic Gaussians Model Srinath Sridhar, Helge Rhodin, Hans-Peter Seidel, Antti Oulasvirta, and Christian Theobalt 3DV 2014 (Oral presentation) [project page] [paper pdf] [video] |
 |
Interactive Motion Mapping for Real-time Character Control Helge Rhodin, James Tompkin, Kwang In Kim, Kiran Varanasi, Hans-Peter Seidel, and Christian Theobalt Eurographics 2014 (Oral presentation) [project page] [paper pdf] [video] |
2013 and before
 |
A Mathematically Justified Algorithm for Shape from Texture
Helge Rhodin and Michael Breuß SSVM 2013 |
 |
A PTX Code Generator for LLVM
Helge Rhodin B.Sc. Thesis, Saarland University, 2010. [project page] [thesis pdf] |
 |
AnySL
The goal of AnySL is the development of a unified shading system that is independent of source language, target architecture and rendering engine without sacrificing runtime performance. My focus was on the GPU code generation. The AnySL project is joined work of the Compiler Design Lab of Prof. Sebastian Hack and the Computer Graphics Group of Prof. Philipp Slusallek. |
 |
Roboking Competition
As Team Hamburg we were able to win twice the international Roboking competition hosted by the Chemnitz University of Technology. The goal was to construct and program an autonomous robot capable of fulfilling the given objectives without human intervention. The team was mentored by Andreas Rhodin and consisted of Julian Bende, Paul Bröker, Daniel Reck, and Helge Rhodin. |
Teaching
2018
 |
EPFL CS 328: Numerical Methods for Visual Computing Visual computing disciplines are characterized by their universal reliance on numerical algorithms to process and display large amounts of visual information such as geometry, images, and volume data. This course will familiarize students with a range of essential numerical tools to solve practical problems in a visual computing context. By Wenzel Jakob and Helge Rhodin |
2017
| |
EPFL CS 328: Numerical Methods for Visual Computing Visual computing disciplines are characterized by their universal reliance on numerical algorithms to process and display large amounts of visual information such as geometry, images, and volume data. This course will familiarize students with a range of essential numerical tools to solve practical problems in a visual computing context. By Wenzel Jakob and Helge Rhodin |
2016
 |
MPI / Saarland University: Computer Vision for Computer Graphics In this seminar series, we will cover advanced research topics that cross the boundaries between the fields of computer vision and computer graphics. We will concentrate on recent research results that were published at top-tier conferences and journals. By Christian Theobalt, Christian Richardt and Dan Casas. Assisted by Helge Rhodin (student supervision, in-class discussion and grading). |
2015
 |
MPI / Saarland University: Computer Vision for Computer Graphics In this seminar series, we will cover advanced research topics that cross the boundaries between the fields of computer vision and computer graphics. We will concentrate on recent research results that were published at top-tier conferences and journals. By Christian Theobalt and Christian Richardt. Assisted by Helge Rhodin (student supervision, in-class discussion and grading). |