Reproduce “UniForge3D: Joint 3D Reconstruction and Generation in a Shared Anchor-Guided Space” 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
Wenjue Chen, Wenyue Chen, Heliang Zheng, Rongfei Jia, Yuan Liu, Meng Li
UniForge3D jointly reconstructs and generates 3D objects from four unposed images by placing both branches in an anchor-guided coordinate space and selecting a supporting view for each voxel. On GSO, it outperforms Mix3R in RGB-rendering PSNR (24.826 versus 21.040) and Chamfer Distance (1.4047 versus 2.3202 ×10⁻²). The benchmark uses synthetic training and does not establish robustness to difficult real-world conditions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.