Welcome!

I'm Thomas Probst, a computer vision researcher based in Zurich, Switzerland.

About

I am a Computer Vision Researcher at Magic Leap.

Previously, I was leading the computer vision research at Specta.ai, a digital healthcare startup. I did my PhD and worked as a Postdoc at the Computer Vision Lab under Prof. Luc van Gool at ETH Zürich (Switzerland). The main focus of my research is on deep learning for geometry problems, and the perception of humans for robotics.

I was working on the EurEyeCase project on robot-assisted eye surgery. I have been involved in the EU research project ReMeDi, aiming to develop a system for tele-operated medical diagnostics.

Before that, I worked on multi-modal image registration under the supervision of Christian Wojek and Heiko Neumann at Carl Zeiss Corporate Research and Technology in Oberkochen (Germany) for my masters thesis.

During my course of studies I engaged in thermal spray coating research and estimation of gas flow velocities at Daimler Research and Development in Ulm (Germany).

I took both my undergraduate and master studies in electrical engineering and computer science at Ulm University.

Selected Publications

teaser
Task Switching Network for Multi-task Learning
Reference Task Switching Network for Multi-task Learning
Guolei Sun, Thomas Probst, Danda Pani Paudel, Nikola Popovic, Menelaos Kanakis, Jagruti Patel, Dengxin Dai, Luc Van Gool
International Conference on Computer Vision (ICCV) 2021
Online, October 2021
teaser
CompositeTasking: Understanding Images by Spatial Composition of Tasks
Reference CompositeTasking: Understanding Images by Spatial Composition of Tasks
Nikola Popovic, Danda Pani Paudel, Thomas Probst, Guolei Sun, Luc Van Gool
Computer Vision and Pattern Recognition (CVPR) 2021
Online, June 2021
teaser
Unsupervised Learning of Consensus Maximization for 3D Vision Problems
Reference Unsupervised Learning of Consensus Maximization for 3D Vision Problems
Thomas Probst, Danda Pani Paudel, Ajad Chhatkuli, Luc Van Gool
Computer Vision and Pattern Recognition (CVPR) 2019
Long Beach, LA, USA
teaser
What Correspondences Reveal About Unknown Camera and Motion Models?
Reference What Correspondences Reveal About Unknown Camera and Motion Models?
Thomas Probst, Ajad Chhatkuli, Danda Pani Paudel, Luc Van Gool
Computer Vision and Pattern Recognition (CVPR) 2019
Long Beach, LA, USA
teaser
Model-free Consensus Maximization for Non-Rigid Shapes
Reference Model-free Consensus Maximization for Non-Rigid Shapes
Thomas Probst, Ajad Chhatkuli, Danda Pani Paudel, Luc Van Gool
European Conference on Computer Vision (ECCV) 2018
Munich, Germany, September 2018

Publications by Topic

3D Vision

Unsupervised Monocular Depth Reconstruction of Non-Rigid Scenes

Ayca Takmaz, Danda Pani Paudel, Thomas Probst, Ajad Chhatkuli, Martin Oswald, Luc Van Gool
3DV 2021
 

Convex Relaxations for Consensus and Non-Minimal Problems in 3D Vision

Thomas Probst, Danda Pani Paudel, Ajad Chhatkuli, Luc Van Gool
International Conference on Computer Vision (ICCV) 2019

Incremental Non-Rigid Structure-from-Motion with Unknown Focal Length
 

Thomas Probst, Danda Pani Paudel, Ajad Chhatkuli, Luc Van Gool
European Conference on Computer Vision (ECCV) 2018
 
 

Automatic Tool Landmark Detection for Stereo Vision in Robot-Assisted Retinal Surgery

Thomas Probst, Kevis-Kokitsi Maninis, Ajad Chhatkuli, Mouloud Ourak, Emmanuel Vander Poorten, Luc Van Gool
International Conference on Robotics and Automation (ICRA) 2018

Looking at Humans

Boosting Crowd Counting with Transformers
 

Guolei Sun, Yun Liu, Thomas Probst, Danda Pani Paudel, Nikola Popovic, Luc Van Gool
arxiv 2021

Dual Grid Net: hand mesh vertex regression from single depth maps

Chengde Wan, Thomas Probst, Luc Van Gool, Angela Yao
European Conference on Computer Vision (ECCV) 2020
 

Self-Supervised 3D Hand Pose Estimation Through Training by Fitting

Chengde Wan, Thomas Probst, Luc Van Gool, Angela Yao
Computer Vision and Pattern Recognition (CVPR) 2019

Dense 3D Regression for Hand Pose Estimation
 

Chengde Wan, Thomas Probst, Luc Van Gool, Angela Yao
Computer Vision and Pattern Recognition (CVPR) 2018

Efficient Model-free Anthropometry from Depth Data
 

Thomas Probst, Andrea Fossati, Mathieu Salzmann and Luc Van Gool
International Conference on 3D Vision (3DV) 2017

Deep Learning on Lie Groups for Skeleton-based Action Recognition

Zhiwu Huang, Chengde Wan, Thomas Probst and Luc Van Gool
Computer Vision and Pattern Recognition (CVPR) 2017

Crossing Nets: Combining GANs and VAEs with a Shared Latent Space for Hand Pose Estimation
 

Chengde Wan, Thomas Probst, Luc Van Gool, Angela Yao
Computer Vision and Pattern Recognition (CVPR) 2017

Combining Human Body Shape And Pose Estimation For Robust Upper Body Tracking Using a Depth Sensor

Thomas Probst, Andrea Fossati and Luc Van Gool
Workshop on Assistive Computer Vision and Robotics (ACVR) 2016

Visual Embeddings

Image-based Navigation using Visual Features and Map

Janine Thoma, Danda Pani Paudel, Ajad Chhatkuli, Thomas Probst, Luc Van Gool
Computer Vision and Pattern Recognition (CVPR) 2019

Learning Feature Representations for Look-Alike Images

Ayca Takmaz , Thomas Probst, Danda Pani Paudel, Luc Van Gool
Vision Meets Cogntition Workshop at Computer Vision and Pattern Recognition (CVPR) 2019

Learning Based Multi-Modal Image Registrations
 

Thomas Probst, Christian Wojek, Heiko Neumann
Masters Thesis
Ulm University and Carl Zeiss Corporate Research, 2014

Contact

thomas.tpr.probst@gmail.com

bitbucket.io/probstt

Template by varadbhogayata