Pengfei Zeng

Papers

Total Citations

H-Index

About

Pengfei Zeng is a robotics researcher whose work centers on advancing robotic manipulation, particularly through innovative approaches to grasping. His major contribution lies in rethinking how robotic grasps are represented and predicted. In his influential paper "CPQNet: Contact Points Quality Network for Robotic Grasping" (2022), Zeng introduced a novel paradigm that simplifies grasp parameterization by representing a grasp using only two contact points, rather than the traditional combination of gripper center, rotation angle, and opening width. This streamlined representation enables more efficient and accurate prediction of grasp quality and pose. Though early in his career, his work has already garnered attention, with the CPQNet paper accumulating 2 citations as a foundational idea in the field. Zeng’s research addresses a critical bottleneck in data-driven grasping—the complexity of grasp representation—and offers a more intuitive, contact-centric framework that could enhance real-world robotic dexterity. His approach is particularly notable for its potential to improve grasp success rates in cluttered or unstructured environments, making it a promising direction for future manipulation systems.