Yonghao Dong

Papers

Total Citations

H-Index

About

Yonghao Dong is a rising researcher in computer vision and robotics, whose work focuses on the critical challenge of pedestrian trajectory prediction—a key component for autonomous navigation and human-robot interaction. His most notable contribution, the "Recurrent Aligned Network for Generalized Pedestrian Trajectory Prediction" (2024), tackles the persistent domain shift problem that limits the real-world applicability of trajectory models. Rather than relying on costly fine-tuning with target domain data, Dong's approach introduces a novel alignment mechanism that enables models to generalize across different environments without retraining. This work, already garnering 12 citations shortly after publication, addresses a fundamental bottleneck in deploying prediction systems in unseen scenarios. Dong's research sits at the intersection of deep learning, domain adaptation, and motion forecasting, with implications for safer autonomous vehicles and more responsive robots. By developing methods that reduce the need for scene-specific data, he is helping to bridge the gap between laboratory-trained models and the unpredictable dynamics of real-world pedestrian behavior—a step toward truly robust, generalizable perception systems.