Explainable AI | Mark Edmonds

Extended prior imitation learning work with human-understandable visual interfaces that describe the robot’s haptic network and And-Or Graph.
Showed that robots may need to “think” one way to complete a task but another way to explain their behavior: the explanations that best fostered trust were not the model components that best helped the robot achieve the task.
Result: treat explainability as a first-class citizen when building AI systems that interact with humans.

Collaborators: Feng Gao, Hangxin Liu, Xu Xie, Siyuan Qi, Brandon Rothrock (NASA JPL), Yixin Zhu, Ying Nian Wu, Hongjing Lu, Song-Chun Zhu

Publications

Sci. Robot.

A tale of two explanations: Enhancing human trust by explaining robot behavior

M. Edmonds*, F. Gao*, H. Liu*, X. Xie*, S. Qi, B. Rothrock, Y. Zhu, Y. N. Wu, H. Lu, and S.-C. Zhu

Science Robotics, 2019

Supplement Article Project page Video

BibTeX

@article{edmonds2019tale,
  title = {A tale of two explanations: Enhancing human trust by explaining robot behavior},
  author = {Edmonds, M. and Gao, F. and Liu, H. and Xie, X. and Qi, S. and Rothrock, B. and Zhu, Y. and Wu, Y. N. and Lu, H. and Zhu, S.-C.},
  journal = {Science Robotics},
  volume = {4},
  number = {37},
  year = {2019},
  doi = {10.1126/scirobotics.aay4663}
}

Sci. Robot.

In situ bidirectional human-robot value alignment

L. Yuan*, X. Gao*, Z. Zheng*, M. Edmonds, Y. N. Wu, F. Rossano, H. Lu, Y. Zhu, and S.-C. Zhu

Science Robotics, 2022

Supplement Article Project page Video

BibTeX

@article{yuan2022situ,
  title = {In situ bidirectional human-robot value alignment},
  author = {Yuan, L. and Gao, X. and Zheng, Z. and Edmonds, M. and Wu, Y. N. and Rossano, F. and Lu, H. and Zhu, Y. and Zhu, S.-C.},
  journal = {Science Robotics},
  volume = {7},
  number = {68},
  year = {2022},
  doi = {10.1126/scirobotics.abm4183}
}