Shengyu Huang

9

Papers

38

Total Citations

Papers (9)

Video Depth without Video Models

Living Scenes: Multi-object Relocalization and Reconstruction in Changing 3D Environments

Rectified Point Flow: Generic Point Cloud Pose Estimation

Dynamic LiDAR Re-simulation using Compositional Neural Fields

Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation

Neural Fields Meet Explicit Geometric Representations for Inverse Rendering of Urban Scenes

Neural LiDAR Fields for Novel View Synthesis

Dynamic 3D Scene Analysis by Point Cloud Accumulation

Predator: Registration of 3D Point Clouds With Low Overlap