Hye-Jung Yoon

DA-Fusion achieves superior segmentation accuracy in cluttered environments by seamlessly integrating RGB and depth data, outperforming traditional methods that struggle with unseen objects.

Seg2Grasp achieves unprecedented adaptability in suction grasping, outperforming traditional methods in complex bin picking scenarios.

Achieving globally consistent 3D instance labels without any 3D-specific training could revolutionize how robots perceive and interact with unknown environments.