Shing Yan Loo

Universiti Putra Malaysia, University of Alberta

Papers

Total Citations

H-Index

About

Shing Yan Loo is a rising researcher in the intersection of computer vision and robotics, specializing in visual simultaneous localization and mapping (VSLAM) and polarimetric imaging. His work addresses critical challenges in dense mapping and depth estimation for autonomous navigation. Loo’s most cited paper, “Polarimetric Monocular Dense Mapping Using Relative Deep Depth Prior” (2021, 16 citations), pioneers the use of polarization cameras to extract surface normal information—azimuth and zenith angles—enhancing dense map reconstruction with relative depth priors. This innovation bridges the gap between photometric and geometric cues. In “Rumination Meets VSLAM: You Do Not Need to Build All the Submaps in Realtime” (2023, 8 citations), he proposes a submap-based VSLAM approach that optimizes computational efficiency by avoiding real-time construction of all submaps, improving long-term tracking robustness. His work “Online Mutual Adaptation of Deep Depth Prediction and Visual SLAM” (2021, 4 citations) explores adaptive integration of deep neural networks for depth prediction, enabling real-time mutual refinement between depth estimation and camera tracking. Loo’s contributions advance practical SLAM applications, from autonomous drones to augmented reality, by fusing deep learning with classical geometric methods. His research is foundational for next-generation visual navigation systems.