VGGT: Visual Geometry Grounded Transformer
π³: Permutation-Equivariant Visual Geometry Learning
G²VLM: Geometry Grounded Vision Language Model with Unified 3D Reconstruction and Spatial Reasoning
Depth Anything 3: Recovering the Visual Space from Any Views
MoRE: 3D Visual Geometry Reconstruction Meets Mixture-of-Experts
Efficiently Reconstructing Dynamic Scenes One D4RT at a Time