machine learning and robotics: bridging the gap between machines and peopleIn the last few years, how robots operate in the world has advanced considerably. Examples include the autonomous vehicles in the DARPA Grand Challenges and Urban Challenge, the considerable work in robot mapping, and the growing interest in home and service robots. However, these example technologies and systems are still mostly restricted to research prototypes. One obstacle to getting more widely useful robots is that the way robots reason about their world is still pretty different to how people reason. Robots think in terms of point features, dense occupancy grids and action cost maps. People think in terms of landmarks, segmented objects and tasks (among other representations). There are good reasons why these are different, and robots are unlikely to ever reason about the world in the same way that people do. But, there has been recent work in bridging the gap between low-level geometry and control, and higher-level semantic representations. I will talk about how machine learning is being used to develop more capable robots that can operate in populated environments and perform complex tasks. I will discuss the state of the art, what the open challenges are and the potential impact of solving these challenges.
Nicholas Roy is an Associate Professor in the Department of Aeronautics & Astronautics at the Massachusetts Institute of Technology
and a member of the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT. He received his Ph. D. in Robotics from
Carnegie Mellon University in 2003. His research interests include autonomous systems, micro air vehicles, mobile robotics,
human-computer interaction, decision-making under uncertainty and machine learning.
machine learning at yahoo!As a large internet portal with diverse audience, Yahoo! is constantly faced with choosing the most appropriate object (a news story, a set of urls, an advertisement) for a specific user in context. It must choose the object(s) out of a large number of choices (millions to billions) in a short time (20-50 milliseconds). Extreme personalization means modeling the user's needs at that moment well. Serving systems must adapt to changes in the environment - e.g shifts in advertising marketplace, drifts in user's needs and tastes, hot news stories becoming cold, etc. The underlying learning machinery must sometimes learn in an adversarial setting (eg email spam detection). While the training data is web-scale that grows terabytes a day, it is still sparse because only a tiny fraction of all possible combinations of webpages, users, and advertisements will be seen in the historical data. A rich combination of mathematics, statistics, computer science, and economics forms the basis of many serving systems at Yahoo. This talk gives an overview of some of the learning/optimization challenges and examines one aspect in depth.
Kishore Papineni graduated with a PhD in Electrical Engineering specializing in feedback control theory from Rice
University in 1995. From 1995 to 2006, he was a Research Staff Member at the IBM T.J. Watson Research Center, Yorktown Heights,
New York. During this period, he worked on various natural language processing technologies such as natural language understanding,
dialog management and statistical machine translation, managing IBM's SMT department from 2001-2006. He was a founding
Editor-in-Chief of ACM Transactions on Speech and Language Processing from 2003-2007. In 2006, he joined Yahoo! Research
where he is the head of Machine Learning. His interests include optimization, estimation, control, and computational advertising.
learning to see peopleThe ability to recognize humans and their activities by vision is key for a machine to interact intelligently and effortlessly with a human-inhabited environment. This talk covers recent research at Daimler R&D and the Univ. of Amsterdam on the topic of "Looking at People"; I cover applications in the intelligent vehicle and smart surveillance domains, and emphasize the central role that machine learning plays herein.
Dariu M. Gavrila received the PhD degree in computer science from the University of Maryland at College Park in 1996. He was a visiting researcher at the MIT Media Laboratory in 1996. Since 1997, he has been a Senior Research Scientist at Daimler R&D in Ulm, Germany. In 2003, he was further named professor at the University of Amsterdam, chairing the area of Intelligent Perception Systems (part time).
Over the last decade, Prof. Gavrila has focused on visual systems
for detecting humans and their activity, with application
to intelligent vehicles and surveillance. His contributions are
frequently cited, and he received the I/O 2007 Award from
the Netherlands Organisation for Scientific Research (NWO).
what is intelligence?
While the question in the title has remained unanswered for thousands of years, it is perhaps easier to address the apparently similar question: "What is intelligence for?" We take a pragmatic approach to intelligent behavior, and we examine systems that can pursue goals in their environment, using information gathered from it in order to make useful decisions, autonomously and robustly. We review the fundamental aspects of their behavior, methods to model it and architectures to realize it. The discussion will cover both natural and artificial systems, ranging from single cells to software agents.
Nello Cristianini is a Professor of Artificial Intelligence at the University of Bristol since March 2006, and a holder of the Royal Society Wolfson Merit Award. He has wide research interests in the area of computational pattern analysis and its application to problems ranging from genomics, to computational linguistics and artificial intelligence systems. He has contributed extensively to the field of kernel methods. He has a PhD from the University of Bristol, a MSc from Royal Holloway, University of London, and a Degree in Physics from University of Trieste. Since 2001 has been Action Editor of the Journal of Machine Learning Research (JMLR), and since 2005 also Associate Editor of the Journal of Artificial Intelligence Research (JAIR). He is co-author of the books 'An Introduction to Support Vector Machines' and 'Kernel Methods for Pattern Analysis' with John Shawe-Taylor, and 'Introduction to Computational Genomics' with Matt Hahn (all published by Cambridge University Press).