Reproduce “QuadLink: Autoregressive Quad-Dominant Mesh Generation via Point-Relation Learning” 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
Yiheng Zhang, Zhe Zhu, Tingrui Shen, Zhuojiang Cai, Tianxiao Li, Zixing Zhao, Qiujie Dong, Zhiyang Dou and 6 more
QuadLink reconstructs point clouds as quad-dominant meshes by predicting vertices and face centroids, learning centroid-conditioned vertex links, and assembling faces, with triangles retained when needed. On 100 models, it reports 0.021 Chamfer Distance and 0.714 Edge Flow Ratio, versus QuadWild’s 0.022 and 0.547, while its quadrilateral ratio is lower: 79.4% versus 100%. These results do not guarantee topology or polygon-count control.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.