Weijia Zeng

Papers

Total Citations

H-Index

About

Weijia Zeng is pioneering the intersection of real-time robot navigation and cutting-edge 3D scene representations. His primary research focuses on developing safe, robust navigation systems for autonomous robots operating in complex environments. His most notable contribution, Splat-Nav, introduces a complete navigation pipeline for Gaussian splatting (GSplat) maps—a powerful emerging 3D scene representation. This work comprises two key modules: Splat-Plan, which enables safe path planning, and Splat-Loc, a vision-based pose estimator that provides robust localization. Already garnering 22 citations since its 2025 publication, Splat-Nav addresses a critical gap by making GSplat scenes practically usable for real-time robotic applications. Zeng’s work is particularly significant because it bridges the gap between advanced computer vision representations and practical robotics, enabling robots to navigate safely through photorealistic 3D reconstructions. His research promises to unlock new capabilities for autonomous systems in applications ranging from indoor service robots to autonomous vehicles, where accurate, real-time navigation through complex environments is essential.