VE2VF: Vision-Enabled to Vision-Free Distillation via Real-world Reinforcement Learning for Robust Contact-Rich Manipulation

Name: VE2VF: Vision-Enabled to Vision-Free Distillation via Real-world Reinforcement Learning for Robust Contact-Rich Manipulation
Uploaded: 2026-10-06T22:17:14.682927+00:00
Duration: 334 s
Description: The VE2VF framework uses a vision-enabled teacher policy trained with human input to distill its knowledge into a vision-free student policy. This allows the student to perform robust contact-rich manipulation using only proprioceptive data like pose and force, achieving strong generalization without needing extensive visual training or augmentation.

The VE2VF framework uses a vision-enabled teacher policy trained with human input to distill its knowledge into a vision-free student policy. This allows the student to perform robust contact-rich manipulation using only proprioceptive data like pose and force, achieving strong generalization without needing extensive visual training or augmentation.

Open the full episode