Jiayao Shan

Northeastern University, Science and Technology on Surface Physics and Chemistry Laboratory

Papers

Total Citations

152

H-Index

About

Jiayao Shan is a researcher specializing in 3D object tracking and LiDAR-based perception, with a particular focus on advancing autonomous driving and robotics systems. Their most influential contribution is the development of the Point-Track-Transformer (PTT) module, a novel architecture designed for 3D single object tracking in point clouds. By integrating transformer-based mechanisms — including feature embedding, position encoding, and self-attention computation — PTT addresses longstanding challenges in extracting meaningful spatial features from sparse and irregular point cloud data. Building on this foundation, Shan extended their work to real-time 3D single object tracking, tackling one of the field's most persistent difficulties: the ambiguity caused by sparse or partially occluded point clouds from objects at long distances, a critical concern in practical autonomous driving scenarios. This follow-up work has garnered 59 citations, reflecting strong community interest in robust, deployable solutions. Collectively, Shan's papers have accumulated over 150 citations, demonstrating meaningful impact within the computer vision and robotics communities. Their research bridges cutting-edge deep learning methodology with real-world applicability, making them a notable emerging voice in LiDAR perception and transformer-based 3D scene understanding.