Yu-Jie Xiong

Oulu Deaconess Institute, Shanghai University of Engineering Science

Papers

Total Citations

H-Index

About

Yu-Jie Xiong’s research spans robotics, sensor design, and cutting-edge deep learning, with a particular focus on point cloud analysis and vision-language navigation. His foundational work introduced the concept of isotropy for robot force sensors, leveraging Fisher’s information matrix to define maximally informative sensor configurations—a contribution that remains cited for its theoretical elegance and practical relevance in robotic manipulation. More recently, Xiong has advanced multimodal AI with a knowledge-distilled pre-training model for vision-language navigation, bridging visual perception and natural language instructions. His 2024 paper on PointABM integrates Bidirectional Mamba and multi-head self-attention for point cloud analysis, challenging Transformer dominance with linear-complexity alternatives. While his earlier work has garnered 17 citations, his newer publications signal growing impact in efficient 3D understanding. Xiong’s trajectory—from sensor isotropy to state-space models—reflects a career dedicated to both foundational theory and modern AI architectures, making his research valuable for students exploring robotics, sensor design, or efficient deep learning.