1 paper · 2 filters
Julian Zilly, Jacopo Tani, Breandan Considine +14
Despite recent breakthroughs, the ability of deep learning and reinforcement learning to outperform traditional approaches to control physically embodied robotic agents remains lar…