1 code implementation • 28 Sep 2021 • Astrid Merckling, Nicolas Perrin-Gilbert, Alex Coninx, Stéphane Doncieux
Our experimental results show that the approach leads to efficient exploration in challenging environments with image observations, and to state representations that significantly accelerate learning in RL tasks.
no code implementations • 15 Sep 2019 • Astrid Merckling, Alexandre Coninx, Loic Cressot, Stéphane Doncieux, Nicolas Perrin-Gilbert
Indeed, a compact representation of such a state is beneficial to help robots grasp onto their environment for interacting.