Changze Li
Papers
1
Total Citations
12
H-Index
1
About
Changze Li is a leading researcher in multimodal sensor fusion and real-time 3D scene understanding, with a focus on integrating LiDAR, inertial, and visual data for robust odometry and mapping. Their most notable contribution, "GS-LIVO: Real-Time LiDAR, Inertial, and Visual Multisensor Fused Odometry With Gaussian Mapping" (2025, 12 citations), introduces a pioneering framework that leverages 3D Gaussian splatting (3D-GS) for efficient, high-fidelity mapping. This work addresses critical limitations in existing vision-only 3D-GS methods, such as reliance on hand-crafted heuristics for point-cloud densification, poor occlusion handling, and excessive GPU memory and computation demands. By fusing LiDAR, inertial, and visual data in real time, Li’s approach achieves superior accuracy and efficiency, advancing the state of the art in autonomous navigation and robotics. Despite its recent publication, the paper has already garnered significant attention, underscoring its impact. Li’s research bridges the gap between traditional sensor fusion and modern neural rendering, offering scalable solutions for dynamic environments. Their work is essential reading for students and researchers exploring real-time mapping, autonomous systems, or 3D scene representation.
Research Focus
Key Achievements
Top Papers
- 1