Papers
5
Total Citations
132
H-Index
4
About
Siwen Quan is a leading researcher in 3D computer vision and robotics, with a focus on point cloud processing, scene understanding, and autonomous systems. Their work centers on developing robust, efficient algorithms for low-cost sensors, addressing critical challenges in 3D feature representation, registration, and scene completion. Quan’s seminal paper, “Local voxelized structure for 3D binary feature representation and robust registration of point clouds from low-cost sensors” (2018, 100 citations), introduced a novel binary feature descriptor that significantly improved point cloud registration accuracy and efficiency, enabling reliable performance with noisy, low-resolution data from affordable LiDAR and depth cameras. This work, alongside their exploration of multi-view silhouette geometry for distinctive 3D features, has laid a foundation for practical mobile robot perception. More recently, Quan’s ESC-Net (2024, 10 citations) tackles the “triple sparsity” problem in extreme sparse LiDAR point clouds, advancing scene completion to directly support downstream mapping and navigation tasks. Their contributions to semi-supervised 3D object detection and multi-instance registration further demonstrate a commitment to data-efficient, real-world solutions. With over 130 citations, Quan’s research is driving the deployment of robust 3D perception in autonomous driving and mobile robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5