Reproduce “Joint Learning of 3D Shape Retrieval and Deformation” 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
Mikaela Angelina Uy, Vladimir G. Kim, Minhyuk Sung, Noam Aigerman, Siddhartha Chaudhuri, Leonidas Guibas
The authors present a jointly trained system that retrieves and deforms 3D models to fit images or scans. Retrieval learns to favor sources that the part-wise deformer can adjust effectively, while deformation learns on sources retrieval favors. On ShapeNet image-to-mesh chair tests, Chamfer distance is 1.005 × 10⁻² versus 1.216 × 10⁻² for DAR+DF; this measures point-cloud fit, not semantic or mesh quality.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.