MIT Reports to the President 1998-99

ARTIFICIAL INTELLIGENCE LABORATORY

The MIT Artificial Intelligence Laboratory has as its principal intellectual goal the understanding of human intelligence. As a practical matter the AI Lab develops the mathematics and engineering of intelligent systems and artifacts.

The MIT AI Lab has been in continuous existence since 1959, and currently has 24 faculty and senior and principal research scientists. The majority of faculty come from the department of Electrical Engineering and Computer Science, along with some from Brain and Cognitive Science, Mechanical Engineering, and Aeronautical and Astronautical Engineering.

Financial support is provided by the Defense Advanced Research Projects Agency (DARPA), the Office of Naval Research (ONR), the National Aeronautics and Space Administration (NASA), the National Science Foundation (NSF), Nippon Telegraph and Telephone (NTT), Ford Motor Corporation, Yamaha Motor Corporation, the Kaiser Foundation, the EPOCH Foundation (Taiwain), Flex Foot Inc., Alphatech Inc., Amgen Inc., Interval Research, Mitsubishi Research Laboratories, and Microsoft.

Starting in July 1st 1998 the AI Lab entered into a significant research partnership with NTT. The Laboratory for Computer Science also participates in this partnership which is administered at MIT through the AI Lab.

RESEARCH

The research activities of the laboratory are divided into six general areas: vision, robotics and cognitive architectures, new models of computation, learning, information management, and language. Two page research abstracts of 114 individual projects at the lab can be found at http://www.ai.mit.edu/lab/abstracts/1998/. Some of the highlights of the year are as follows.

Prof. Eric Grimson and his students developed more techniques for image guided surgery and continued to support a growth in use of their systems in brain surgery cases at Brighman and Women's Hospital. Prof. Grimson's group also furthered development of their real-time tracking system which monitors multiple cameras 24 hours a day recognizing vehicular and pedestrian traffic. After learning in an entirely unsupervised manner, the system is able to notice anomalous behaviors and report them to a user in real-time.

Prof. Poggio and his students built robust trainable vision systems able to reliably detect pedestrians in static images, including recognizing and classifying their body parts.

Prof. Viola and his students began a new project which uses multiple cameras at fixed positions and angles to look at a common scene. Their aim is to be able to synthesize any viewpoint (perhaps moving, and perhaps related to what is happening in the scene) in real-time from this stream of fixed views.

Professor Gill Pratt and his students continued their development of walking robots but also built a new artificial knee to be used as a human prosthesis. Dr. Ken Salisbury and his students extended their work on haptic virtual reality building a new laparoscopic surgery simulator. Prof. Rodney Brooks and his students completed development of new control methods for compliant robot arms on the humanoid robot Cog, and developed a new robot Kismet that is able to both express itself socially and to pick up on social cues from people around it.

Dr. Thomas Knight and his students fabricated and tested the first fully reversible computer. The system has a 12 bit CPU making it more than a simple conceptual demonstrator–the processor can be run forwards or backwards; possible because it never does the energy-costly step of completely forgetting a bit. Dr. Knight also got a new DNA computing laboratory up and running–the idea is to use living cells to do computation by inserting fabricated pieces of DNA into them and having the transcription mechanism be hijacked into doing the desired computations.

Prof. Tomaso Poggio and his students developed new models of learning object class in the visual cortex. They also significantly extended the theoretical underpinnings of learning on Support Vector Machines (SVMs). Prof. Lynn Stein and her students looked at human learning of computer science and developed new approaches to teaching introductory courses, changing the fundamental model of computation.

A number of students and faculty continued the development of the lab's Intelligent Room, and ported their work to a new Intelligent Office. This required new developments in agent systems, new vision systems, and integration with language understanding systems.

Prof. Davis, Dr. Shrobe, and their students designed new tools for knowledge navigation, knowledge capture, and knowledge maintenance.

Dr. Boris Katz and his students extended their START natural language system to understand more about Web pages and be able to answer more complex queries. They also connected it to a commercial voice recognition system making it possible to browse web content by simply asking questions.

More information about the AI Lab can be found on the World Wide Web at http://www.ai.mit.edu/.

Rodney A. Brooks

MIT Reports to the President 1998-99