UniForge3D Open-Sourced: Reconstructs Full 3D Assets from Sparse Unposed Images

Peking University, Math Magic, and HKUST researchers have released and open-sourced UniForge3D, which jointly performs 3D reconstruction and generation from sparse, unposed images. It models both tasks in a shared anchor-guided space, preserving observed geometry while completing unseen regions, achieving SOTA results.

2026-10-14 ~ 2026-10-15 · 2 related posts

• UniForge3D From PKU and Others Hits SOTA in 3D Asset Generation From Sparse Pose-Free Images — PekingUniversity · 2026-10-14
• UniForge3D turns sparse-view images into high-fidelity 3D meshes, code released — camenduru · 2026-10-15