Leidi Zhao
Papers
5
Total Citations
33
H-Index
3
About
Leidi Zhao is a robotics researcher whose work sits at the intersection of human-robot interaction, control systems, and crowdsourced machine learning. Her primary research focuses on enabling robots to acquire and synthesize complex physical skills—from assistive movements to dexterous manipulation—by learning from diverse human demonstrations. A standout contribution is her integrated framework for robotic Sit-To-Stand assistance, which combines control design with human intention recognition to create more responsive and natural assistive robots (22 citations). Zhao has also pioneered approaches to crowdsourced robot learning, developing state space discretization methods that allow robots to learn from large, heterogeneous datasets provided by multiple human mentors. This work breaks the traditional limits of learning from a single demonstrator, enabling robots to synthesize entirely new skills—such as the nunchaku flipping challenge—without extensive retraining. Her research on "Synthesis of Robot Hand Skills Powered by Crowdsourced Learning" and "Handling Crowdsourced Data Using State Space Discretization" has laid important groundwork for scalable, data-driven robot skill acquisition. By merging control theory with data-centric learning, Zhao is helping to create robots that can physically coexist with humans more safely and adaptively.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Synthesis of Robot Hand Skills Powered by Crowdsourced Learning3 citations · 2019
- 4Robot Composite Learning and the Nunchaku Flipping Challenge3 citations · 2018
- 5