home (original) (raw)

Peter Gehler

Is trying to learn how to learn to see.

Curriculum Vitae Google Scholar LinkedIn ResearchGate dblp

Welcome!

I am a researcher at Amazon in Tübingen. Before, I was W3 professor at the University of Würzburg, group leader at the Bernstein Center of Computational Neuroscience at the University of Tübingen and affiliated with the Max Planck Institute for Intelligent Systems in Tübingen as a senior research scientist.

I want to enable computers to understand the physical world around us. I am mostly interested in models that infer semantic and physical properties of visual data.

We have several openings for internship on a number of topics, year round. Please drop me an email if you are interested.

Postdocs

• Raghudeep Gadde , now Amazon

Students

• Sergey Prokudin , now ETH
• Christoph Lassner , now Amazon Meta Reality Labs
• Thomas Nestmeyer , now Hyundai
• Varun Jampani , now Nvidia Research Google Research
• Martin Kiefel , now Google Amazon
• Andreas Lehrmann , now Disney Research, Facebook Reality Labs Borealis AI

Publications

	The Role of Pretrained Representations for the OOD Generalization of RL Agents Andrea Dittadi , Frederik Traeuble , Manuel Wüthrich , Felix Widmaier , Peter Gehler , Ole Winther , Francesco Locatello , Olivier Bachem , Bernhard Schölkopf , and Stefan Bauer International Conference on Learning Representations (ICLR), 2022 pdf arxiv

University Courses

• Intelligent Systems – Graphical Models in Computer Vision Universität Tübingen, 4h/week course, winter term 2014/2015 website
• Intelligent Systems – Graphical Models in Computer Vision Universität Tübingen, 4h/week course, summer term 2013 website
• Probabilistic Graphical Models and their Applications Saarland University, 4h/week course, winter term 2011/2012, with Bernt Schiele website
• Machine Learning 1 TU Darmstadt, 4h/week course, summer term 2011 website
• Machine Learning 2 TU Darmstadt, 5h/week course, winter term 2010/2011 website

Tutorials and Practicals

• Lecture - Research Network on Learning Systems 2014, Zurich Structured Output Models in Computer Vision website
• Practical on Kernel Methods - MLSS Machine Learning Summer School A practical session on kernel methods at the MLSS 2013 in Tübingen. This is a programming exercise on the use of kernel methods and parameter selection. website
• Tutorial - ENS/INRIA Summer School 2013, Paris Introduction to Graphical Models slides website

Chair Positions and Memberships

Member of ELLIS.

• Area Chair - CVPR, 2020
• Area Chair - ECCV, 2014, 2016, 2020
• Area Chair - ICCV, 2015, 2019
• Area Chair - NeurIPS, 2016, 2018, 2019
• Area Chair - ICML, 2017, 2018
• Program Chair - German Conference on Pattern Recognition, 2015
• Associate Editor for IEEE Transactions on Pattern Analysis and Machine Intelligence - TPAMI

Conference and Workshop Organization

• Inverse Rendering - ICCV, 2015, link
• 37th German Conference on Pattern Recognition - GCPR, 2015, link
• Structured Prediction - Tractability, Learning and Inference - CVPR, 2013, link
• At the intersection of Vision, Graphics, Learning and Sensing - Representations and Applications - Cambridge, UK, 2012, link
• 1st IEEE Workshop on Kernels and Distances for Computer Vision - ICCV - 2011, link
• Stuctured Models in Computer Vision, CVPR, 2010, link

Mountainbiking

• 10 Things I learned about mountainbiking from the great Filme von Draussen