Park Kunbum

Papers

Total Citations

H-Index

About

Park Kunbum is a researcher whose work lies at the intersection of computer vision, robotics, and spatial intelligence. His primary research focuses on 3D reconstruction, visual-inertial odometry (VIO), and the integration of pretrained deep learning features for robust environmental perception. In his most-cited paper, "3D Reconstruction by Pretrained Features and Visual-Inertial Odometry" (2022), Kunbum proposed a novel framework that leverages pretrained neural network features to enhance the accuracy and resilience of 3D mapping in dynamic or texture-poor environments. This work bridges the gap between classical geometric methods and modern learning-based approaches, offering a practical solution for autonomous navigation and augmented reality systems. While his citation count is still growing, the paper’s methodological clarity and potential for real-world deployment have already attracted attention from peers in the field. Kunbum’s contributions are particularly notable for their emphasis on combining sensor fusion with deep learning, a direction that promises to advance the reliability of spatial AI. His ongoing work continues to explore how pretrained models can reduce computational overhead while maintaining high-fidelity reconstruction, marking him as a promising voice in the next generation of 3D vision researchers.