…Each image is annotated with keypoints, pixel-to-vertex correspondences (from CSE ) and a segmentation mask.
2 PAPERS • NO BENCHMARKS YET