Xia Li

Papers

Total Citations

H-Index

About

Xia Li is an emerging researcher at the intersection of computer vision, multimodal learning, and 3D scene understanding, with a particular focus on advancing perception systems for real-world autonomous applications. Their most notable work, "VG4D: Vision-Language Model Goes 4D Video Recognition" (2024), represents a pioneering contribution to the field of 4D point cloud video understanding — a domain critical to robotics and autonomous driving. In this work, Li addresses a fundamental challenge in the field: the inherent limitations of sensor resolution that deprive existing methods of fine-grained spatial and temporal detail. By bridging the power of vision-language models with 4D video recognition, Li's approach leverages rich semantic knowledge from large-scale multimodal pretraining to compensate for sparse point cloud data — a creative and impactful methodological leap. While still in its early citation trajectory with 5 citations since publication, the work has already attracted attention from the robotics and autonomous systems communities. Li's research signals a promising trajectory in developing robust, language-guided perception systems capable of interpreting dynamic 3D environments with greater accuracy and contextual awareness.