Sangni Xu

Papers

Total Citations

H-Index

About

Sangni Xu is a researcher whose work sits at the intersection of computer vision and robotics, with a particular focus on visual odometry (VO)—the process of estimating a camera's position and trajectory from sequential images. In her highly regarded 2021 paper, "Attention-based Long-term Modeling for Deep Visual Odometry," Xu introduced a novel deep learning framework that leverages attention mechanisms to capture long-range dependencies in image sequences. This approach addresses a critical limitation of conventional VO methods, which often rely on hand-crafted features and struggle with drift over time. By integrating attention-based modeling, her work significantly improves the accuracy and robustness of VO systems, making them more reliable for real-world applications such as autonomous driving, augmented/virtual reality, and robotics. Although her most-cited paper currently holds six citations—a modest number reflecting the early stage of her career—the innovative nature of her contribution positions her as a promising voice in the field. Xu’s research bridges the gap between traditional geometric methods and modern deep learning, offering a path toward more intelligent and adaptive visual perception systems.