A new large-scale dataset that consists of 409 fine-grained categories and 31,881 images with accurate 3D pose annotation.
1 PAPER • NO BENCHMARKS YET