Pengjie Xu
Papers
15
Total Citations
123
H-Index
5
About
Pengjie Xu is a robotics researcher whose work spans deep reinforcement learning, visual place recognition, multi-robot coordination, and autonomous navigation. His research addresses some of the most demanding challenges in modern robotics: enabling machines to operate reliably in complex, uncertain, and collaborative environments. Xu's most celebrated contribution — a task-adaptive deep reinforcement learning framework for dual-arm robot manipulation (46 citations) — tackles the inherently difficult problem of closed-chain manipulation, where coordinated robot arms must execute flexible, adaptive tasks together. Complementing this, his reinforcement learning-based distributed impedance control approach advances compliant robot operation in tight interaction scenarios, pushing beyond the limitations of traditional model-based controllers. In visual perception, Xu has pioneered robust place recognition systems for mobile robots, including a knowledge distillation framework using soft-hard label teaching (20 citations) and a dark-enhanced network that extends recognition capabilities into low-light environments — a critical step toward real-world deployment. His additional contributions in multi-robot SLAM map fusion, zero-shot hierarchical path planning, and cooperative transportation with mobile manipulators reveal a researcher committed to building holistic, deployable robotic systems. With over 100 cumulative citations largely from 2024 publications alone, Xu is rapidly establishing himself as a rising force in intelligent robotics research.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6Map Fusion Method Based on Image Stitching for Multi-robot SLAM5 citations · 2021
- 7
- 8
- 9
- 10