Yuan Rao

Papers

Total Citations

H-Index

About

Yuan Rao is a leading researcher in underwater robotics and 3D vision, specializing in photometric stereo reconstruction, visual feature matching, and semantic-aware localization. His most impactful work, "Underwater Surface Normal Reconstruction via Cross-Grained Photometric Stereo Transformer" (2024, 23 citations), introduces a novel transformer-based method for recovering high-precision 3D surface normals of textureless underwater objects like seabeds and pipelines—a critical capability for autonomous underwater robots. This contribution directly addresses the challenge of acquiring dense, accurate 3D data in degraded underwater environments. Rao also developed "Underwater visual feature matching based on attenuation invariance" (2023, 3 citations), which tackles the problem of feature descriptor instability caused by light attenuation, color distortion, and blur in underwater images, enhancing robot localization and navigation. His recent work, "Learning Semantic-Aware Point-Line Features for Localization and Reconstruction" (2025, 1 citation), advances beyond traditional point features to incorporate line features, improving robustness in marine engineering detection and autonomous navigation. With a focus on overcoming the unique visual challenges of underwater environments, Rao’s research is foundational for next-generation underwater robotics and 3D mapping systems.