Xiaoxiao Long

Papers

Total Citations

H-Index

About

Xiaoxiao Long is a computer vision and robotics researcher whose work sits at the intersection of 3D scene understanding, depth estimation, and language-guided robot manipulation. Perhaps best known for the Epipolar Spatio-Temporal Networks framework for multi-view depth estimation, Long developed a novel approach to inferring consistent depth maps from single video sequences — a technically demanding challenge with direct applications in autonomous perception, 3D reconstruction, and robot navigation. This work, which accumulated over 55 citations, advanced the field by addressing temporal consistency limitations that had hampered earlier learning-based methods. More recently, Long has pushed into the emerging domain of 3D Gaussian Splatting, contributing GaussianGrasper, a system that constructs richly queryable 3D scene representations to enable open-vocabulary robotic grasping based on natural language directives. Attracting 37 citations shortly after its 2024 publication, this work reflects a timely fusion of neural rendering and embodied AI — two of the most rapidly evolving frontiers in the field. Together, Long's contributions demonstrate a consistent focus on making 3D spatial understanding more robust, generalizable, and practically deployable in real-world robotic systems.