Computer science graduate student receives AI accolades

Department of Computer Science and Engineering graduate student Timothy Mann was recently chosen to participate in the National Science Foundation's East Asia and Pacific Summer Institutes program (NSF EAPSI), and the Association of the Advancement of Artificial Intelligence Doctoral Consortium (AAAI DC).

Photo of Timothy MannBoth programs are highly selective and their invitations are a reflection of the high quality of Mann's research.

NSF EAPSI fosters collaboration between US graduate students and East Asian and Pacific universities by providing an all expenses paid summer visit to conduct research at these institutions. Mann was chosen to work with Dr. Minho Lee at Kyungpook National University in Daegu, South Korea. Mann's research will focus on improving near and far stereo-visual depth estimation of objects.

"Dr. Lee's group has developed a method of using two cameras to emulate stereoscopic vision in order to judge depth, but it is only accurate when the target object is within a certain range of distance. What I hope to do is expand that range of accuracy by introducing a robot that will interact with the camera's target object to learn that object's physical properties; properties such as size and shape, and incorporate the robots information with the camera's information to produce a more accurate representation of depth," Mann said.

Mann will conduct his research for eight weeks while in South Korea.

Tim was also recently invited to attend the prestigious AAAI Doctoral Consortium. The consortium kicks off the yearly AAAI conference, and allows the top PhD students in Artificial Intelligence (AI) to interact with other top students, scholars, and professionals within this field. Tim will be presenting a summary of his dissertation research, which focuses on designing reinforcement learning algorithms that efficiently learn complex motor behaviors using transfer learning (i.e. applying information gained while learning simple skills to more complex problems). By starting with simple problems a learning system can incrementally learn more advanced skills--each step applying what it has learned previously.

"This is a great opportunity to meet with the leaders in AI, and get their feedback and guidance regarding my research within the field. It will also provide direction for any future research and perhaps foster career paths that may form."

Tim is a Ph.D. candidate whose research interests include robotics, machine learning, and reinforcement learning. He is advised by Associate Professor of Computer Science and Engineering Dr. Yoonsuck Choe.