We are interested in developing and enhancing methods that allow robots to adapt to specific users and personalize their interaction in order to provide long-term effective human-robot interaction for socially assistive robotics as well as other domains.

As robots are expected to perform in increasingly complex domains, the ability to automatically adapt and learn has become more critical. Interaction with people, a single user or different individuals/users, is an especially challenging learning problem, as it involves a "moving target", since people's mood, health state, task expertise, and other factors that are not necessarily accessible to the robot, change at various time-scales and can influence behavior in unexpected ways.

Human-robot interaction requires a unique perspective on the learning process. Traditional machine learning techniques typically require extensive data, but people typically do not have the patience to provide ample training examples, nor do they always act consistently. Finally, people's behavior changes over time, on multiple timescales. Additionally, people require a natural interface. Thus, learning and adaptation in HRI must find a balance between power and robustness of algorithms, allowing for and leveraging having people in the loop, and adapting quickly enough for human tolerance and with limited data. Our work on learning is driven by several motivations:

• To leverage human-robot interaction in order to enhance the robot's ability to learn new concepts, skills, and tasks
• To enable the robot to personalize its interaction with the user so as to optimize its effectiveness in assisting the user
• To develop methods for learning at human timescales, which are both very short (within a social interaction) and very long (repeated encounters/sessions, months, and even years)
• To situate HRI learning in realistic domains, including classrooms, health care contexts, and homes

Our ongoing projects are exploring learning and adaptation in a variety of problem domains, including human-robot team collaboration, user adaptation, task learning from demonstration, and socially guided service task learning. Our past work as also studied imitation learning in depth, as described here., and our ongoing work is building on those results. The project list below points to more details about each of these efforts.