Interested? Read on, please.
My research focuses on the computational and perceptual aspects of human-robot interaction. Specifically, I look at developing algorithms that would enable mobile robots to collaborate safely and robustly in the field with humans. The word "field" is loosely used to describe an environment that has not been specifically designed or engineered for mobile robots, and which includes the physical space occupied by humans in their personal or professional life. This includes underwater, aerial, and terrestrial domains, where robots are needed to operate in all kinds of terrain conditions alongside, and/or under human guidance. Communication, perception, dialog, and learning are core elements of my research, in addition to systems design, both software, and hardware.
Certain issues including but not limited to the psychological impacts of robots on humans and ethics in robotics are important components of human-robot interaction; however, students interested in pursuing research in such areas are likely to find other groups better suited to their interest.
Also, while we do a fair amount of research in robot vision, we are primarily a robotics group. Students who I work with have a passion for robots, field robots in particular, and real-time computing and robot development. Students are expected to contribute to the advancement of robotics. Our research problems start as a concept and end up as implemented, working algorithms running onboard physical robots in realistic environments. If you are mainly interested in computer vision and deep learning, we probably are not the best fit for you.
Students interested in pursuing graduate studies as part of my group should have a keen interest in probabilistic robotics, robot perception, machine learning, and/or interaction design. Good coding skills are necessary as the research will be validated on actual robotic platforms, often outdoors and in challenging environments. Familiarity with hardware (Arduinos, embedded boards, actuators, and general robot hardware and sensors) is a plus.
Of course, this is all subject to openings in the lab, which may not always exist.