Yanhui Duan
Papers
2
Total Citations
40
H-Index
2
About
Yanhui Duan is a leading researcher in robotic manipulation and computer vision, with a focus on bridging the sim-to-real gap for autonomous grasping systems. Her work centers on two critical challenges: enabling robots to understand object affordances from partial 3D data, and transferring learned skills from simulation to the real world without costly manual labeling. Her most influential paper, "Grasp Pose Detection with Affordance-based Task Constraint Learning in Single-view Point Clouds" (2020, 33 citations), introduced a novel framework that learns task-specific constraints from affordances, allowing robots to generate stable grasps from a single point cloud view—a significant advance for real-time applications. In her 2023 work on pixel-level domain adaptation for object pose estimation (7 citations), Duan tackles the persistent domain gap problem by enforcing cycle-consistency and content preservation between synthetic and real images, dramatically improving transfer accuracy. Her contributions are particularly notable for reducing the cost and complexity of deploying robotic systems in unstructured environments, making her work essential reading for researchers in manipulation, domain adaptation, and embodied AI.
Research Focus
Key Achievements
Top Papers
- 1
- 2Pixel-Level Domain Adaptation for Real-to-Sim Object Pose Estimation7 citations · 2023