Jianzhi Lyu

Universität Hamburg

Papers

3

Total Citations

56

H-Index

3

About

Jianzhi Lyu is a roboticist whose research lies at the intersection of manipulation, perception, and human-robot interaction. His work focuses on enabling robots to perform complex tasks in unstructured environments, with key contributions in reinforcement learning for dexterous manipulation and domain adaptation for robust perception. Lyu is perhaps best known for his innovative approach to grasping objects from "ungraspable" poses—such as a book lying flat on a table—by teaching a robot to first push the object to the edge and then grasp it from the overhanging portion, a strategy inspired by human manipulation. This work, published in 2023, has already garnered 25 citations, highlighting its immediate impact. He has also advanced unsupervised domain adaptation through a novel method combining hypothesis transfer with gradual knowledge distillation (19 citations), enabling robots to adapt their perception to changing environments without retraining. Additionally, Lyu has explored natural language grounding for human-robot interaction, developing systems that parse referring expressions and scene graphs to resolve ambiguity in commands (12 citations). His research is paving the way toward more intuitive, adaptable, and capable robotic systems.

Research Focus

Key Achievements

3
H-Index
3
Papers
56
Total Citations
19
Avg Citations/Paper
🏆 Most Cited Paper
Reinforcement Learning Based Pushing and Grasping Objects from Ungraspable Poses
25 citations · 2023
📈 Most Prolific Year: 2023 (1 Papers)
🤝 Key Collaborators: 14
🏛 Institutions: Universität Hamburg

Top Papers

  1. 1
  2. 2
  3. 3

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 12 days ago