Search papers, labs, and topics across Lattice.
1
0
2
0
By constraining reward functions to Control Barrier Functions, this approach achieves safe adversarial imitation learning that adapts directly from expert observations without requiring labeled data.
