Xiangyun Meng

University of Washington

Papers

8

Total Citations

105

H-Index

4

About

Xiangyun Meng is a roboticist whose research spans visual topological navigation, dexterous manipulation, and agile legged locomotion. His most influential work, “Scaling Local Control to Large-Scale Topological Navigation” (2020, 61 citations), revitalized visual navigation by leveraging deep learning to overcome scalability and reliability challenges in real-world environments. In manipulation, Meng’s “Hierarchical Policies for Cluttered-Scene Grasping With Latent Plans” (2022, 22 citations) addresses the longstanding problem of 6D grasping in cluttered scenes, proposing a hierarchical framework that outperforms open-loop pipelines and end-to-end methods. For legged robots, he introduced CAJun (2023), a hierarchical learning and control framework enabling continuous, adaptive jumping—a breakthrough for dynamic locomotion in discontinuous terrains like stairs and stepping stones. Meng also pioneered semantics-aware locomotion skills from human demonstration (2022), allowing quadrupeds to adjust behaviors based on terrain type. His work on composable behavior embeddings for long-horizon visual navigation (2021) further advances robot autonomy by enabling compact visual memory and sparse topological planning. With over 100 total citations and contributions spanning perception, planning, and control, Meng’s research is shaping the next generation of autonomous robots capable of navigating, grasping, and moving with unprecedented agility and reliability.

Research Focus

Key Achievements

4
H-Index
8
Papers
105
Total Citations
13
Avg Citations/Paper
🏆 Most Cited Paper
Scaling Local Control to Large-Scale Topological Navigation
61 citations · 2020
📈 Most Prolific Year: 2022 (2 Papers)
🤝 Key Collaborators: 16
🏛 Institutions: University of Washington

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago