Haofei Kuang
Papers
6
Total Citations
95
H-Index
5
About
Haofei Kuang is a robotics researcher whose work sits at the intersection of localization, mapping, and 3D perception. His primary contributions lie in developing novel approaches for robot pose estimation and environment mapping, particularly using implicit neural representations. Kuang’s most influential work, "IR-MCL," introduces an implicit representation-based approach to Monte Carlo localization, achieving 32 citations by enabling more accurate global localization from 2D LiDAR data. His closely related "LocNDF" paper (31 citations) advances neural distance field mapping specifically optimized for robot localization tasks, demonstrating how modern neural fields can outperform traditional occupancy maps. Kuang has also made notable contributions to visual odometry, rethinking the Fourier-Mellin Transform for multi-depth camera views, and to underwater robotics, proposing an unsupervised method for depth estimation from spherical images. His earlier work on fast Gaussian Process Occupancy Maps addressed critical computational bottlenecks in real-time mapping. Across his publications, Kuang consistently tackles the challenge of making sophisticated geometric and learning-based methods practical for real-world robotic systems, with his research accumulating over 95 citations and establishing him as a rising figure in robot perception and localization.
Research Focus
Key Achievements
Top Papers
- 1IR-MCL: Implicit Representation-Based Online Global Localization32 citations · 2023
- 2LocNDF: Neural Distance Field Mapping for Robot Localization31 citations · 2023
- 3
- 4Fast Gaussian Process Occupancy Maps9 citations · 2018
- 5Underwater Depth Estimation for Spherical Images7 citations · 2021
- 6Rotation Estimation for Omni-directional Cameras Using Sinusoid Fitting4 citations · 2021