Haiyong Jiang

University of Chinese Academy of Sciences

Papers

Total Citations

H-Index

About

Haiyong Jiang is a leading researcher in 3D computer vision and geometric deep learning, with a focus on point cloud understanding and scene perception. His work addresses critical challenges in robotics and augmented reality, particularly in 3D instance segmentation and point cloud assembly. Jiang’s most cited paper, "End-to-End 3D Point Cloud Instance Segmentation Without Detection" (2020, 38 citations), introduces a groundbreaking method that bypasses traditional detection branches or grouping steps, enabling direct instance segmentation from raw point clouds—a significant advancement for real-time environment perception. This work has influenced subsequent research in autonomous navigation and AR. More recently, his paper "PuzzleNet: Boundary-Aware Feature Matching for Non-Overlapping 3D Point Clouds Assembly" (2023, 7 citations) tackles the challenging task of assembling fragmented 3D scans without prior alignment, using boundary-aware feature matching to achieve robust reconstruction. Jiang’s contributions are notable for their practical impact on robotics and AR, where efficient and accurate 3D understanding is critical. His research continues to push the boundaries of geometric learning, offering innovative solutions for complex 3D data processing.