Peidong Liu

Papers

Total Citations

209

H-Index

About

Peidong Liu is a leading researcher in robotics and computer vision, specializing in 3D mapping, visual odometry, and aerial manipulation. His seminal work on robust dense mapping for large-scale dynamic environments (2018, 147 citations) introduced a pioneering stereo-based algorithm that simultaneously reconstructs static backgrounds, moving objects, and stationary but potentially mobile elements—a critical advancement for autonomous navigation in complex urban settings. Liu further advanced direct visual odometry for fisheye-stereo cameras (2017, 49 citations), enabling precise motion estimation and semi-dense reconstruction in challenging, wide-angle environments. His recent contributions include robust localization of occluded targets in aerial manipulation via range-only mapping (2022, 7 citations), addressing the practical challenge of continuous target tracking when onboard manipulators obstruct vision. Most notably, Liu’s latest work, IncEventGS (2025, 6 citations), achieves pose-free Gaussian splatting from a single event camera, pushing the boundaries of bio-inspired visual sensing for novel view synthesis. With a strong focus on real-world deployment, Liu’s research bridges theoretical innovation and practical robotics, earning him recognition as a key figure in dynamic scene understanding and autonomous systems.