Search papers, labs, and topics across Lattice.
University of California Berkeley United States
1
0
30
STL-shaped rewards lead to tighter velocity tracking and more stable training for quadruped locomotion, outperforming traditional hand-crafted methods.
