The IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR) is one of the three premier international conferences in the field of pattern recognition and computer vision. This year this conference took place in Salt Lake City, US. As part of the program, leading academic institutions (Stanford, ETH Zurich, Max-Planck Institute for Intelligent Systems in Tubingen, University of Heidelberg) organized a Robust Vision Challenge with the purpose of identifying the world's most accurate algorithms for solving different tasks in the area of computer vision. In this context, the main goal was to develop algorithms that are robust, i.e. that work equally well for a variety of different data sets.
Daniel Maurer, PhD student at the Institute for Visualization and Interactive Systems, took part in this challenge and with a recently developed algorithm. With this algorithm he achieved an outstanding second place in the field of motion estimation for which he was finally awarded the CVPR 2018 Robust Vision Challenge Runner-Up Award at the corresponding conference workshop.