Papers
3
Total Citations
38
H-Index
3
About
Hengcan Shi is a computer vision and robotics researcher whose work centers on 3D perception, autonomous systems, and scene understanding in complex real-world environments. His research addresses some of the most demanding challenges in robotic vision, particularly the accurate detection and tracking of pedestrians from 3D point cloud data — a critical capability for autonomous driving and human-robot interaction. Shi's most recognized contribution is his development of the Efficient Attentive Pillar Network, a framework designed for accurate and real-time 3D pedestrian detection that tackles the inherent difficulties of human body pose variation and sparse LiDAR point clouds. This work has accumulated 31 citations, reflecting its significance to the autonomous driving and robotics communities. His research extends beyond detection into broader scene understanding, as demonstrated by his role in creating JRDB-PanoTrack, an open-world panoptic segmentation and tracking dataset tailored for crowded human environments — a resource that supports multi-sensor robot navigation and human-robot interaction research. Collectively, Shi's contributions push the boundaries of real-time perception and multi-modal sensing, making him a notable emerging voice in robotic vision research.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3