Search papers, labs, and topics across Lattice.
1
0
3
Vision-free robotic manipulation policies can be trained in under an hour to achieve 95% success and generalize to unseen tasks by distilling knowledge from vision-enabled teachers using real-world reinforcement learning.
