Hongpeng Cao

Papers

Total Citations

H-Index

About

Hongpeng Cao is a robotics researcher whose work sits at the intersection of computer vision, deep reinforcement learning, and safe AI for autonomous manipulation. His most impactful contribution is **6IMPOSE**, a framework designed to bridge the "reality gap" in 6D pose estimation for robotic grasping. While deep learning models excel on benchmarks, Cao’s work directly addresses their poor generalization in real-world settings—a critical step for practical deployment. This flagship paper has already garnered **18 citations** since 2023, signaling strong interest from the manipulation community. In parallel, Cao tackles high-precision industrial tasks, such as flexible gear assembly, by fusing YOLO-based coarse localization with deep reinforcement learning for fine insertion, achieving both speed and accuracy. His research also extends to safety assurance, where he models DNN-based controllers in stochastic games to formally verify their robustness—a vital concern for deploying AI in safety-critical systems. By combining rigorous engineering with formal methods, Cao is shaping a future where robots can grasp, assemble, and operate reliably outside the lab.