Naoya Sogi
Papers
2
Total Citations
21
H-Index
2
About
Naoya Sogi is a researcher whose work sits at the intersection of computer vision and robotics, with a particular focus on planar pose estimation (PPE) and robust geometric perception. His key research area addresses the fundamental challenge of accurately estimating the 6DOF pose of planar markers from single images—a critical task for augmented reality, mapping, and localization. Sogi’s major contribution lies in resolving the inherent ambiguity in marker pose estimation, which arises from the symmetry and planar nature of markers. He introduced a novel method that leverages robust rotation averaging with clique constraints, enabling the disambiguation of multiple plausible poses and yielding more reliable and accurate estimates. This work, published in 2020, has garnered 18 citations, reflecting its practical importance in the field. By tackling a long-standing problem in marker-based tracking, Sogi’s approach enhances the robustness of systems that depend on precise camera-to-marker alignment. His research is particularly valuable for students and engineers working on visual SLAM, drone navigation, or industrial inspection, where marker-based localization must be both fast and unambiguous.
Research Focus
Key Achievements
Top Papers
- 1
- 2