Filters








1 Hit in 1.8 sec

ORStereo: Occlusion-Aware Recurrent Stereo Matching for 4K-Resolution Images [article]

Yaoyu Hu, Wenshan Wang, Huai Yu, Weikun Zhen, Sebastian Scherer
2021 arXiv   pre-print
In this work, we present the Occlusion-aware Recurrent binocular Stereo matching (ORStereo), which deals with these issues by only training on available low disparity range stereo images.  ...  Compared to other methods that are only trained on low-resolution images, our method is 70% more accurate on 4K-resolution images.  ...  Fig. 1 . 1 ORStereo: Occlusion-aware Recurrent binocular Stereo matching. The whole model is trained with low-resolution data. During high-resolution testing, ORStereo in used in two phases.  ... 
arXiv:2103.07798v1 fatcat:crvos256ejfs7pbwtq2d677z6i