Haocheng Peng
Papers
2
Total Citations
8
H-Index
2
About
Haocheng Peng is a robotics researcher pushing the boundaries of embodied AI through innovative approaches to motion planning and visual localization. His work centers on two critical challenges: enabling robots to navigate complex physical environments and achieving robust global localization at scale. Peng’s flagship paper, "PC-Planner," introduces a physics-constrained self-supervised learning framework for neural motion planning, leveraging a shape-aware distance function to overcome the high-dimensional complexities that plague traditional methods. This work has already garnered 5 citations since its 2024 publication, signaling its early impact on the field. In parallel, his research on satellite-assisted visual localization tackles the persistent drift errors in long-term SLAM by fusing satellite imagery with ground-level semantic matching—a novel approach that has earned 3 citations. By bridging global satellite views with local robot perception, Peng is helping to solve one of visual SLAM’s most stubborn problems: cumulative tracking error in large-scale environments. His contributions are particularly timely as the robotics community races toward robust, real-world deployment of autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2