(1) Single-view Geometry: Single-view Depth/Normal/Boundary Estimation; Single-view Object/Scene Reconstruction (2) Two-view/N-view Geometry: Structure from Motion; N-view 3D Reconstruction (3) Neural Geometric Processing: Point Cloud/Mesh/Voxel Learning; Neural Implicit Functions for 3D Representation (4) Neural Rendering: Differentiable Volumetric Rendering; Neural Scene Representation and Application (5) Generative 3D Models: 3D GANs; 3D Diffusion Models (6) Dynamic 3D Models: Scene Flow Estimation; Dynamic 3D Reconstruction. Students should seek the course instructor’s approval to take this course.