Learned Collaborative Stereo Refinement

Patrick Knöbelreiter, Thomas Pock

Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-review


In this work, we propose a learning-based method to denoise and refine disparity maps of a given stereo method. The proposed variational network arises naturally from unrolling the iterates of a proximal gradient method applied to a variational energy defined in a joint disparity, color, and confidence image space. Our method allows to learn a robust collaborative regularizer leveraging the joint statistics of the color image, the confidence map and the disparity map. Due to the variational structure of our method, the individual steps can be easily visualized, thus enabling interpretability of the method. We can therefore provide interesting insights into how our method refines and denoises disparity maps. The efficiency of our method is demonstrated by the publicly available stereo benchmarks Middlebury 2014 and Kitti 2015.
Original languageEnglish
Title of host publicationGerman Conference on Pattern Recognition
Publication statusPublished - 2019
Event41th German Conference on Pattern Recognition - Dortmund, Germany
Duration: 10 Sept 201913 Sept 2019


Conference41th German Conference on Pattern Recognition
Abbreviated titleGCPR 2019
Internet address

Cite this