FoldNet: Learning Generalizable Closed-Loop Policy for Garment Folding via Keypoint-Driven Asset and Demonstration Synthesis
Accepted in RAL, 2026
Recommended citation: Y. Chen, B. Xiao and H. Wang, "FoldNet: Learning Generalizable Closed-Loop Policy for Garment Folding via Keypoint-Driven Asset and Demonstration Synthesis," in IEEE Robotics and Automation Letters, vol. 11, no. 3, pp. 3820-3827, March 2026, doi: 10.1109/LRA.2026.3656770. https://ieeexplore.ieee.org/document/11359673
We builds a pipeline for automatic cloth generation and trained a diffusion policy for cloth folding using just synthetic data.