About the Talk “What do we want robots to do for us (and how do we get there)?”: With the great advances in robotics from the last years, we are at the point to ask ourselves “what do we really want robots to do for us”? Answering this question could bring a human component into robotics and AI research early on. In this talk, I will talk about our efforts to find out what people want robots to do for them, and how can we enable research towards that direction. I will then present some of my lab’s projects to bring us closer to that goal in two ways: creating solutions to understand human instructions, and models to predict future images based on agent’s actions.
About the Speaker: Roberto Martin-Martin is an Assistant Professor of Computer Science at the University of Texas at Austin and director of the Robotic Interactive Intelligence Lab. His research integrates robotics, computer vision, and machine learning. He is interested on integrating physical interaction as synergistic component of robot perception, planning and manipulation skills. Previously, he worked as Postdoctoral scholar at the Stanford Vision and Learning Lab with Professors Silvio Savarese and Fei-Fei Li, where he coordinated research projects in two groups: the People, AI & Robots (PAIR) team, studying visuo-motor learning skills for manipulation and planning, and the iGibson/BEHAVIOR team, examining problems in long-horizon mobile manipulation and human cognition. He also led the Jackrabbot team, where he studied mobile manipulation in human-populated environments. He received his doctoral and master’s degree from Technische Universität Berlin (TUB), and his BSc. degree from Universidad Politécnica de Madrid. Roberto’s work has been awarded the RSS Best System Paper Award and nominee for ICRA Best Robot System Paper Award, IROS Best Paper Award, Best Student Paper Award, and Best Cognitive Science Paper Award. He has been selected RSS Pioneer, and he is chair of IEEE/RAS Technical Committee in Mobile Manipulation.


GIPHY App Key not set. Please check settings