Yan Zha
Papers
6
Total Citations
62
H-Index
4
About
Yan Zha’s research lies at the intersection of human-robot interaction, explainable AI, and robot learning, with a focus on making autonomous agents more intuitive and aligned with human expectations. Her most influential work introduces the concept of “explicable planning,” which formalizes how an AI agent can minimize the distance between its planned behavior and what a human would naturally expect—a foundational idea for trustworthy human-AI collaboration. This line of research, published in 2016 and 2019, has garnered over 50 citations and is widely recognized for addressing the critical gap between autonomous capability and human interpretability. Zha has also advanced robot learning from demonstration, notably through a contrastive visual attention method that uses affordance cues for robotic grasping, and through the development of NatSGD, a multimodal dataset combining speech, gestures, and demonstrations for natural human-robot interaction. Her recent work tackles the challenge of learning from ambiguous demonstrations by integrating self-explanation into reinforcement learning, pushing the boundaries of how robots can learn effectively from imperfect human input. With contributions spanning explicability, multimodal interaction, and robust learning, Zha is shaping the future of robots that can understand and be understood by people.
Research Focus
Key Achievements
Top Papers
- 1Explicable Planning as Minimizing Distance from Expected Behavior29 citations · 2019
- 2Explicable Robot Planning as Minimizing Distance from Expected Behavior.20 citations · 2016
- 3
- 4
- 5
- 6Explicablility as Minimizing Distance from Expected Behavior2 citations · 2016