Jijunnan Li

Papers

Total Citations

H-Index

About

Jijunnan Li is a computer vision researcher whose work centers on visual localization, camera re-localization, and scene understanding — foundational challenges in enabling machines to perceive and navigate the physical world with precision. His research addresses the critical problem of determining accurate six-degree-of-freedom (6DoF) camera poses within pre-established 3D environments, with direct applications in augmented reality, robotics, and autonomous driving. Li's most influential contribution, "Pose Refinement with Joint Optimization of Visual Points and Lines" (2022, 18 citations), advances beyond conventional point-based re-localization by incorporating line features, significantly improving robustness in textureless or feature-sparse environments where earlier methods fall short. His integrated framework RLOCS, presented in "Retrieval and Localization with Observation Constraints" (2021), combines image retrieval and semantic understanding to deliver more reliable end-to-end visual re-localization. Further demonstrating his commitment to practical deployment, his real-time fusion framework tackles persistent real-world challenges including motion blur, illumination shifts, and environmental variations. Across his body of work, Li consistently bridges theoretical innovation with real-world applicability, making meaningful strides toward robust, deployable localization systems that can perform reliably under demanding and unpredictable conditions.