“Knowledge-based Computer Graphics: Intersections of AI and Computer Graphics” Moderated by Marek Holynski

  • ©Alan Borning, Steven K. Feiner, Mark Friedell, and Fanya S. Montalvo



Entry Number: 10


    Knowledge-based Computer Graphics: Intersections of AI and Computer Graphics



Additional Information:

    As graphic interaction becomes more and more complex, present computer-graphics technology for managing such interaction becomes increasingly inadequate. Knowledge-based graphics systems, using domain-specific knowledge, may be the key to powerful new techniques for interaction management. Knowledge-based graphics systems tie the meaning of a picture to its graphics representation and, through these underlying relationships, support the user in deciding about graphics presentation. These systems use artificial intelligence techniques that help to organize and manipulate knowledge specific to graphics in the form of semantic networks, rules and constraints.
    This panel reviews and discusses some of the promising new directions that are beginning to emerge in this exciting area. This panel considers the following topics: recent work on constraint-based graphics; rules for determining the objects to be shown in a picture; matching levels of abstraction to the user’s need to control detail; graphical definitions of new kinds of constraints and the use of constraints in a “viewing and filtering” model of user interfaces; a general framework for the graphic input problem, i.e., how to recognize potentially ambiguous input; problems of reference in graphic interaction; adaptive graphics interfaces; application of inductive learning techniques for an effective visual representation; inference reasoning for scene composition.

Overview Page: