Seongmin Lee

Papers

Total Citations

H-Index

About

Seongmin Lee is a researcher at the intersection of computer vision and robotics, with a primary focus on advancing camera calibration and 3D reconstruction techniques. His most-cited work, "Deep Chessboard Corner Detection Using Multi-task Learning" (2021, 4 citations), tackles a foundational challenge in these fields: the accurate detection of matching correspondences essential for understanding spatial structure. By leveraging multi-task learning, Lee’s approach improves the robustness of chessboard corner detection—a critical step for applications ranging from augmented reality to camera motion estimation. This contribution addresses a persistent bottleneck in automated calibration pipelines, offering a learning-based solution that outperforms traditional methods in challenging real-world conditions. Lee’s research underscores the growing importance of deep learning in geometric computer vision, bridging the gap between classical calibration algorithms and modern neural architectures. His work is particularly valuable for students and engineers developing autonomous systems, where precise calibration directly impacts performance. With a clear focus on practical, implementation-ready solutions, Lee continues to contribute to the foundational tools that enable more reliable and accurate visual perception in robotics and mixed reality environments.