23 June 2022 | Tübingen/New Orleans

Andreas Geiger wins Longuet-Higgins prize

Second high-profile award for the “KITTI Vision Benchmark”

Andreas Geiger, professor of computer vision and machine learning at the University of Tübingen, is part of a team of researchers honored with the prestigious Longuet-Higgins prize at this year’s IEEE Conference on Computer Vision and Pattern Recognition (CVPR). In addition to Geiger, Philip Lenz and Raquel Urtasun received the prestigious award for their joint work on the “KITTI Vision Benchmark,” which they have been running and developing since 2012.

“It is fantastic to see how KITTI has been driving research in computer vision over the last ten years as recognized by this award”, said Geiger. “When we started out, we were not sure if all the hard work would pay off eventually. It was a risky project that has continuously evolved with new benchmarks and data. It is still actively maintained by myself and many international collaborators that contributed time and dedication.”

This is the second high-profile award for Geiger and his team – in October 2021, the researchers were honored with the Everingham Prize for KITTI. It’s also the second time in three years, that the Longuet-Higgins prize is awarded to a scientist from the Cyber Valley ecosystem: In 2020, Michael J. Black, Director at the Max Planck Institute for Intelligent Systems, won it along with his former students Deqing Sun and Stefan Roth for their paper “Secrets of optical flow estimation and their principles”.

KITTI is among the world’s most popular datasets used in the development of self-driving cars and related machine vision applications. The dataset established new tasks and fostered new research in the area of perception for self-driving vehicles, in particular in stereo, flow, scene flow, monodepth, odometry, detection, tracking and semantic segmentation. The KITTI benchmark has over 24,500 registered users, has received more than 28,500 submissions since its inception and currently lists 1905 public entries in its various leaderboards. The KITTI benchmarks has led to various follow-up datasets, benchmarks and workshops in the computer vision community.

The original paper “Are we ready for Autonomous Driving? The KITTI Vision Benchmark Suite” by Geiger, Lenz and Urtasun was published at CVPR in 2012. Now, ten years later, it is recognized by this award due to its significant impact on computer vision research in the past decade. The prize is named after theoretical chemist and cognitive scientist H. Christopher Longuet-Higgins. The award is presented by IEEE Pattern Analysis and Machine Intelligence (PAMI) Technical Committee at each year’s CVPR. Winners are decided by a committee appointed by the TCPAMI Awards Committee.