Welcome to our website
We develop computer vision methods, including 3D object detection, hand pose estimation, and geo-localization, with application to augmented reality and robotics.
- We are co-organizing the Second International Workshop on Recovering 6D Object Pose at ECCV'16.
Our CDL for Semantic 3D Computer Vision opened on April 26!
Coverage (in German): on the CDG website, Krone, Futurezone.
- Prof. Lepetit is a Program Chair for ACCV'16.
July 29, 2016. Two papers accepted at ECCV'16:
- LIFT: Learned Invariant Feature Transform (Spotlight), and
- Going Further with Point Pair Features.
- July 1st, 2016. Invited talk at the 2nd HANDS workshop at CVPR'16.
- June 7, 2016. Invited talk at Google Zurich.
- May 20, 2016. Invited talk at the ImVis workshop organized by CAMP.
- April 29, 2016. One paper accepted at MICCAI'16: Age Estimation from Hand MR Volumes using Deep Learning.
March 4, 2016. Three papers accepted at CVPR'16:
- Learning to Assign Orientations to Feature Points (Oral),
- Direct Prediction of 3D Body Poses from Motion Compensated Sequences, and
- Efficiently Creating 3D Training Data for Fine Hand Pose Estimation.
- February 3rd, 2015. Invited talk at the WARM'16 workshop at TU Graz.
- February 1st, 2015. Invited talk at the Engineering Department of the University of Cambridge.
- December 17, 2015. We co-organized the 1st International Workshop on Recovering 6D Object Pose, in conjunction with ICCV'15.
- November 23, 2015. Invited talk at the DICTA 2015 conference, Adelaide, Australia.
- October 1, 2015. Our paper on geolocalization from 2D maps and images received the Best Paper award at ISMAR'15.
- September 10, 2015. Invited talk at Ecole Centrale de Paris.
August 29, 2015. Four papers accepted at ICCV'15:
- Training a Feedback Loop for Hand Pose Estimation (Oral),
- A Novel Representation of Parts for Accurate 3D Object Detection and Tracking in Monocular Images,
- Projection onto the Manifold of Elongated Structures for Accurate Extraction, and
- An Efficient Minimal Solution for Multi-Camera Motion.