Universität Bielefeld - Sonderforschungsbereich 360
Incremental Evaluation and Simultaneous Natural Language Description
of Time-varying Scenes
Gerhard Herzog, Universität des Saarlandes
Abstract:
The interplay between vision and language in human-machine-interaction
constitutes a prominent issue in many potential application areas and
thus increasingly attracts attention in current research. A major goal
of language-oriented AI research in this context is to attain a completely
operational form of referential semantics that reaches down to the sensoric
level. In this presentation, we will report on practical experience gained
in the project VITRA (VIsual TRAnslator) of the SFB 314, special collaborative
program on AI and knowledge-based systems. The aim of our approach towards
the integration of computer vision and natural language processing is to
obtain an incremental evaluation and simultaneous description of real world
image sequences. The talk provides a detailed description of the VITRA
conception and presents results of our experimental investigations. Current
limitations of the approach and future research tasks will be discussed as
well.
Anke Weinberger, 1995-10-18