Dinh-Cuong Hoang
FPT University, National Taiwan University, National Taipei University of Technology
Papers
11
Total Citations
149
H-Index
7
About
Dinh-Cuong Hoang is a computer vision and robotics researcher whose work sits at the intersection of 3D perception, pose estimation, and robotic manipulation. With a research trajectory spanning nearly a decade, he has made significant contributions to how machines understand and interact with the physical world. Hoang's earliest recognized work, dating to 2016, established novel methodologies for multi-view 3D point cloud registration and robotic object scanning, laying a technical foundation that would inform his later research. His most impactful contribution — "Voting and Attention-Based Pose Relation Learning for Object Pose Estimation From 3D Point Clouds" (2022, 41 citations) — introduced attention mechanisms to the challenging problem of 6DOF object pose estimation, addressing real-world complications such as occlusion and measurement noise. This work has become a reference point in the field. More recently, Hoang has expanded into grasp detection, hand-object interaction, and multimodal fusion, producing a prolific cluster of publications in 2024 alone. His research on graspability-aware pose estimation and collision-free grasp detection reflects a commitment to bridging perception with practical robotic deployment. Collectively accumulating over 140 citations, his body of work represents a coherent and growing contribution to intelligent robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Graspability-Aware Object Pose Estimation in Cluttered Scenes18 citations · 2024
- 4Grasp Configuration Synthesis from 3D Point Clouds with Attention Mechanism17 citations · 2023
- 5
- 6Collision-Free Grasp Detection From Color and Depth Images9 citations · 2024
- 7Attention-based hand pose estimation with voting and dual modalities8 citations · 2024
- 8
- 9
- 10Attention-Based Grasp Detection With Monocular Depth Estimation5 citations · 2024