The research team led by Professor Lee Sung-won from Kookmin University's Department of Electronic Engineering has made a notable achievement by having two papers accepted for presentation at the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026), recognized as the foremost international conference in the field of AI and computer vision. This event is highly regarded due to its rigorous peer-review process and low acceptance rate, where only leading universities and global research institutions present high-caliber research findings.

The joint research, conducted in collaboration with Professor Kim Eun-tae's team from Yonsei University, resulted in significant advancements that integrate theoretical sophistication with practical scalability. The papers address core issues in contemporary technology: the first paper, "Rethinking Pose Refinement in 3D Gaussian Splatting under Pose Prior and Geometric Uncertainty," introduces a novel optimization framework that systematically analyzes and improves camera pose errors and geometric uncertainties, allowing for more stable and precise 3D reconstructions. These advancements underscore the potential applications in diverse fields such as digital twins, robotics, and immersive content creation.

The second paper, titled "HypeVPR: Exploring Hyperbolic Space for Perspective to Equirectangular Visual Place Recognition," delves into the recognition of visual places from different viewpoints and could enhance future developments in machine perception and spatial recognition technologies. Together, these contributions not only highlight the capabilities of Kookmin University's faculty and students but also significantly elevate the institution's reputation in the global academic community, showcasing South Korea's growing prominence in AI research.