Xiang Song

Papers

Total Citations

H-Index

About

Xiang Song is a leading researcher in autonomous systems and 3D perception, with a primary focus on pedestrian trajectory prediction and 3D object detection for self-driving cars and robotics. His most impactful work, “A Novel Graph-Based Trajectory Predictor With Pseudo-Oracle” (2021, 69 citations), addresses the critical challenge of predicting pedestrian motion in dynamic scenes by modeling social interactions and future uncertainty through a graph-based framework—a key contribution to safe autonomous navigation. Song also advanced 3D perception with “Scale-Aware Attention-Based PillarsNet (SAPN)” (2020), which enhances object detection in point clouds by integrating scale-aware attention mechanisms, improving accuracy for applications like autonomous driving and housekeeping robots. His research bridges the gap between social interaction modeling and spatial perception, earning recognition for its practical impact on real-world robotics. With a growing citation record, Song’s work is essential reading for students and researchers tackling trajectory forecasting and LiDAR-based detection, offering innovative solutions that push the boundaries of how machines understand and navigate complex environments.