Learning active manipulation to target shapes with model-free, long-horizon deep reinforcement learning.

Matias Sivertsvik, Kirill Sumskiy, Ekrem Misimi

Browse the full ICRA paper archive.