Kaijian Liu

Zhejiang University, Stevens Institute of Technology

Papers

Total Citations

H-Index

About

Kaijian Liu is a researcher at the forefront of robotic perception and infrastructure inspection, with key contributions in semantic mapping and underwater computer vision. His work bridges the gap between low-level sensor data and high-level robotic cognition, enabling machines to understand and interact with complex environments. Liu’s most cited paper, "Object-aware Semantic Mapping of Indoor Scenes using Octomap" (2019, 6 citations), addresses a critical limitation in robotics: the inability to organize and maintain semantic knowledge of indoor spaces beyond mere 3D reconstruction. By integrating object-level awareness into octree-based mapping, he provides a framework for robots to deliver sophisticated services in cluttered, real-world settings. More recently, Liu has advanced automated infrastructure safety with "Underwater vision-enhanced image segmentation for supporting automated inspection of underwater bridge components" (2025, 2 citations), tackling the challenging domain of turbid water imaging to improve the reliability of robotic bridge inspections. Though early in its impact, this work signals a promising direction for civil infrastructure monitoring. Liu’s research is essential reading for those interested in embodied AI, semantic scene understanding, and the application of computer vision to critical infrastructure maintenance.