7 datasets found

The dataset used for stereo reconstruction, consisting of Middlebury (MB) dataset.

The dataset used for stereo reconstruction, consisting of semi-supervised training set Tr = {(er, e+, e−)n}n=1:N.

The dataset used for stereo reconstruction, consisting of semi-supervised training set Tr = {(er, e+, e−)n}n=1:N.

The OpenSurfaces dataset is a large-scale dataset for material segmentation, containing 25,352 high-resolution indoor images with 45 material categories.

The Local Material Database (LMD) is a dataset for material segmentation, containing 5,845 low-resolution images with 16 mutually exclusive material classes.

The RBI dataset includes more than 10,000 image patches extracted from 400 whole slide images with biopsy-proven results.

A deep learning model that learns subject-level representation from a set of local features. The model represents the image volume as a bag (or set) of local features and can...