the Computer Science Department
AI and Cognition
What computer processes can produce the kind of problem solving, learning, self-awareness, perception, commonsense, communication abilities, and other cognitive functions exhibited by humans? This question drives much of our research in AI and cognition. We study it by building autonomous robotic agents and software agents for the web, observing the behavior of humans and gerbils, constructing general-purpose architectures for intelligent agents, and more.
Faculty
Joe Bates Believable autonomous agents that have the qualities people see in traditional animated characters. Rich personality, the illusion of life, emotion, and thought.
Jaime Carbonell Machine translation, planning and learning.
Jill Fain Natural language and integrated architectures, such as SOAR.
Tom Mitchell Architectures for agents that learn from experience, including robot and web-based agents. Especially interested in the role of machine learning in these architectures.
Jack Mostow Getting computers to listen to children read aloud, and help them. Project LISTEN = "Language Instruction (or Literacy Innovation) that Speech Technology ENables."
Reid Simmons Self-reliant autonomous robots that can plan, monitor their plans, and react to contingencies. Especially interested in robot architectures that integrate reaction and deliberation, probabilistic planning and navigation, and selective perception. Interested in integrating machine learning and robot architectures.
Raul Valdes-Perez Scientific discovery and computers, e.g., programs to carry out scientific reasoning at its highest levels. Practical deployment of tools as scientist's assistants. Implications for future organization of science.
Manuela Veloso Planning and learning by combining inductive and deductive techniques, including analogical/case-based reasoning. Minds for robots: Experience-based agents that plan, execute, and learn. Collaborative and adversarial planning and learning: RoboSoccer and Bolo. Machine learning for signal understanding through evolutionary computation. Rationale capture and reuse in mixed-initiative planning.
previous next csd home maintained by mitchell@cs.cmu.edu