Max Robotics — Global Service Robot Directory + US Certification

About

Xue Bin Peng is a leading researcher at the intersection of deep reinforcement learning, character animation, and legged robotics. His work focuses on developing intelligent controllers that enable both simulated characters and physical robots to move with naturalness, agility, and robustness — bridging the gap between data-driven animation and real-world deployment. Peng's most celebrated contribution is **DeepMimic** (2018), which demonstrated that reinforcement learning could be used to teach physically simulated characters to imitate motion capture data while responding dynamically to perturbations — a landmark result that has accumulated over 800 citations and reshaped the field of physics-based character animation. Building on this foundation, his research expanded into legged robotics, producing highly influential work on bipedal and quadrupedal locomotion control, including robust jumping, versatile walking, and even soccer-playing robots with Cassie and quadrupedal platforms. A recurring theme in his research is **imitation learning from natural sources** — whether human motion capture or animal footage — to produce controllers that generalize robustly to real-world environments. His work on continual fine-tuning of locomotion policies in the real world further highlights his commitment to practical, deployable robotics. With over 1,400 citations across his top papers, Peng has established himself as one of the most impactful young researchers in physically grounded AI and robot learning.