A Text Understanding Program for Modifying Facial Line Drawings on a CRT.

摘要

This report describes a cooperative interaction between human recognition and descriptive powers and a text understanding program for generating facial images. Language design and implementation techniques of the developing program are discussed with special attention to requirements for conversational communication. Excerpts from an example dialogue demonstrate resolution of relational ambiguities and maintenance of contextual information by the interactive program. All processing is done in an IMLAC PDS-1 minicomputer and display processor of 8K 16 bit/word memory. (Author)