Papers
2
Total Citations
27
H-Index
2
About
Xiangting Li is a researcher at the forefront of embedded computer vision, specializing in hardware acceleration and efficient algorithms for Visual Simultaneous Localization and Mapping (VSLAM). Their work directly addresses the critical challenge of deploying advanced neural network-based techniques on resource-constrained mobile and robotic platforms. Li’s major contributions include pioneering the design of FPGA-based hardware accelerators for real-time keypoint extraction, as demonstrated in their highly cited work "MobileSP" (2022, 24 citations), which enables the use of state-of-the-art CNN models like SuperPoint on mobile devices. Furthermore, Li developed "AdaSG" (2022), a lightweight feature point matching method that leverages Graph Neural Networks (GNNs) with adaptive descriptors to significantly improve matching performance in VSLAM systems. By bridging the gap between powerful deep learning algorithms and practical, real-time deployment, Xiangting Li is making VSLAM technology more accessible and robust for autonomous navigation, augmented reality, and robotics, establishing a strong foundation for future innovations in efficient visual perception.
Research Focus
Key Achievements
Top Papers
- 1
- 2