Learning-based approaches to quadruped locomotion commonly adopt generic policy
architectures like fully connected MLPs. As such architectures contain few inductive biases …

We present a comprehensive framework for studying and leveraging morphological
symmetries in robotic systems. These are intrinsic properties of the robot's morphology …

Model-free reinforcement learning is a promising approach for autonomously solving
challenging robotics control problems, but faces exploration difficulty without information of …

Humanoid activities involving sequential contacts are crucial for complex robotic interactions
and operations in the real world and are traditionally solved by model-based motion …

Robot soccer, in its full complexity, poses an unsolved research challenge. Current solutions
heavily rely on engineered heuristic strategies, which lack robustness and adaptability …

This paper proposes a novel method to improve locomotion learning for a single quadruped
robot using multi-agent deep reinforcement learning (MARL). Many existing methods use …