vision.middlebury.edu/stereo (original) (raw)

Welcome to the Middlebury Stereo Vision Page. This website accompanies our taxonomy and comparison of two-frame stereo correspondence algorithms [1], extending our initial paper with Ramin Zabih [2]. It contains:

• An on-line evaluation of current algorithms
• Many stereo datasets with ground-truth disparities
• Our stereo correspondence software
• An on-line submission script that allows you to evaluate your stereo algorithm in our framework

How to cite the materials on this website:
We grant permission to use and publish all images and numerical results on this website. If you report performance results, we request that you cite our paper [1]. Instructions on how to cite our datasets are listed on the datasets page. If you want to cite this website, please use the URL "vision.middlebury.edu/stereo/".

References:

Other online stereo benchmarks:

• Robust Vision Challenge
• KITTI Stereo 2012 evaluation
• KITTI Stereo 2015 evaluation
• ETH3D 2-view stereo benchmark
• Heidelberg HD1K Stereo benchmark

Support for this work was provided in part by NSF CAREER grant 9984485 and NSF grant IIS-0413169. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.