Reproduce “Category-Specific Object Reconstruction from a Single Image” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Abhishek Kar, Shubham Tulsiani, João Carreira, Jitendra Malik

Kar et al.'s system reconstructs 3D shapes and image-aligned depth from photographs by learning deformable category models from 2D annotations, then fitting them to detected objects using segmentation and viewpoint. Mean mesh error was 7.16 versus Carvi's 7.60 in a same-dataset comparison, and 8.24, 8.24, and 10.84 in a held-out setting with ground-truth boxes. Depth refinement was not quantitatively evaluated.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.