ETH Zurich - D-INFK - IVC - CVG - People - PhD Students

PhD Students


ETH Zürich
Universitätstrasse 6
CH-8092 Zürich

Department of Computer Science
Computer Vision and Geometry Group
CNB G 100.9

+41 (0)44 632 96 75

I left ETH and am no longer updating this website. Please find my new website and contact details on


  • June 2016: I finished my PhD and will start as a postdoc at University of California at Berkeley in July
  • June 2016: We are giving a tutorial about semantic 3D modeling at CVPR
    Tutorial material available on
  • February 2016: My plane sweeping stereo matching implemenation is publicly available –>
  • October 2015: We are giving a tutorial about semantic 3D modeling at 3DV –>
    Tutorial material available on
  • July 2015: IROS paper about obstacle detection for self driving cars accepted –>.
  • March 2015: Three papers accepted for CVPR –>.
  • September 2014: Our 3DV submission "Real-Time Direct Dense Matching on Fisheye Images Using Plane-Sweeping Stereo" was accepted as oral.
  • February 2014: Our CVPR submission "Class Specific 3D Object Shape Priors Using Surface Normals" was accepted.
  • September 2013: Our ICCV submission "Real-World Normal Map Capture for Nearly Flat Reflective Surfaces" was accepted.
  • May 2013: I am doing a 3 month internship at Microsoft Research in Cambridge UK.

About Me

From 2011 - 2016 I was working as a PhD student in the Computer Vision and Geometry Group of the Institute for Visual Computing at the Eidgenössische Technische Hochschule Zürich (ETHZ), under the supervision of Prof. Marc Pollefeys.

My research focused on semantic 3D modelling. The goal is to get a dense 3D reconstruction from images and a semantic segmentation of the 3D model at the same time. I also worked on the dense mapping (online and offline) for the V-Charge project, which aims for parking and charging electrical vehicles autonomously.

The video shows the results of our CVPR 2013 paper "Joint 3D Scene Reconstruction and Class Segmentation"

Joint 3D Scene Reconstruction and Class Segmentation

© CVG, ETH Zürich