Robotics and AI: Promises and Challenges
The digitization of practically everything coupled with the mobile Internet, the automation of knowledge work, and advanced robotics promises a future with democratized use of machines and wide-spread use of customization and data-driven decision making. Advances are happening in three different but overlapping fields: robotics, machine learning, and artificial intelligence. However, the science and engineering of intelligence remains a grand challenge.
Where are the gaps we need to address in order to advance toward a future where machines help people reliably with cognitive and physical tasks? In this talk I will discuss recent developments toward enabling robust machine learning and pervasive robotics.
Daniela Rus is the Andrew (1956) and Erna Viterbi Professor of Electrical Engineering and Computer Science; Director of the Computer Science and Artificial Intelligence Laboratory (CSAIL); and Deputy Dean of Research for Schwarzman College of Computing at MIT. Prof. Rus brings deep expertise in robotics, artificial intelligence, data science, and computation. She is a member of the National Academy of Engineering, a member of the American Academy of Arts and Sciences, and fellow of the Association for the Advancement of Artificial Intelligence, the Institute of Electrical and Electronics Engineer, and the Association for Computing Machinery. She is also a recipient of a MacArthur Fellowship, a National Science Foundation Career award, and an Alfred P. Sloan Foundation fellowship. Rus earned her PhD in computer science from Cornell University.
Building Knowledge For AI Agents With Reinforcement Learning
Reinforcement learning allows autonomous agents to learn how to act in a stochastic, unknown environment, with which they can interact. Deep reinforcement learning, in particular, has achieved great success in well-defined application domains, such as Go or chess, in which an agent has to learn how to act and there is a clear success criterion. In this talk, I will focus on the potential role of reinforcement learning as a tool for building knowledge representations in AI agents whose goal is to perform continual learning. I will examine a key concept in reinforcement learning, the value function, and discuss its generalization to support various forms of predictive knowledge. I will also discuss the role of temporally extended actions, and their associated predictive models, in learning procedural knowledge. In order to tame the possible complexity of learning knowledge representations, reinforcement learning agents can use the concepts of intents (ie intended consequences of courses of actions) and affordances (which capture knowlege about where actions can be applied). Finally, I will discuss the challenge of how to evaluate reinforcement learning agents whose goal is not just to control their environment, but also to build knowledge about their world.
Doina Precup splits her time between McGill University/MILA, where she holds a Canada CIFAR AI (CCAI) Chair, and DeepMind Montreal, where she has led the research team since its formation in October 2017. Her research interests are in the areas of reinforcement learning, deep learning, time series analysis, and diverse applications of machine learning in health care, automated control, and other fields. She became a senior member of the Association for the Advancement of Artificial Intelligence in 2015, Canada Research Chair in Machine Learning in 2016, Senior Fellow of the Canadian Institute for Advanced Research in 2017. Dr. Precup is also involved in activities supporting the organization of the wider Montreal, Quebec and Canadian AI ecosystems.
Knorr Moritz - Bosch Corporate Research
Abstract: Modern day camera calibration in computer vision dates back 20 years to a milestone paper from Zhang , which provided instructions along with code to allow novice users to perform monocular camera calibration using only readily available components. His work excelled in terms of usability and robustness and is therefore one of the most cited contributions in the field. Despite this amazing work and the many contributions since then the surge of cameras in mobile devices, robots, and automobiles calls for ever new, more accurate, and more flexible calibration methods.
In this keynote presentation, I would like to explore the many facets of camera calibration, from new approaches in deep learning, which are able to estimate the parameters of a camera model from a single image to highly accurate methods using thousands of parameters, and from single cameras to heterogeneous multi-camera networks.
Finally, I will address one of the main challenges of calibration, which is assessing the calibration result itself.
 Z. Zhang, A flexible new technique for camera calibration
Moritz Knorr is a senior research engineer at Bosch Corporate Research. He has worked both in researching new calibration methods and in product development.
He received his PhD from the Karlsruhe Institute of Technology for his work on self-calibration of multi-camera calibration systems.
His research interests are in the field of single and multi-camera calibration with a focus on enabling non-expert users to perform accurate calibration in diverse scenarios.