UNIVERSITY OF HERTFORDSHIRE COMPUTER SCIENCE RESEARCH COLLOQUIUM presents "Articulate Perception with CHREST: Learning to Identify and Describe Hierarchical Objects in Images speaker: Dr. Peter C. Lane (School of Computer Science, University of Hertfordshire) 5 May 2004 (Wednesday) Lecture Theatre LC108 Hatfield, College Lane Campus 3 - 4 pm Coffee/tea and biscuits will be available. Everyone is Welcome to Attend [Space Permitting] Abstract: CHREST (Chunk Hierarchy and Retrieval Structure) is a computational model of how humans acquire and use perceptual knowledge to work with visual information. Phenomena such as the time to acquire a certain level of expertise, details of eye movements, and the affect of expectations on perception have all been captured by the model. Recent work has extended the symbolic representations used by CHREST to support information from multiple modalities, specifically, images at the level of bitmaps and verbal descriptions of those images. I shall demonstrate and explain the operation of CHREST in learning and retrieving information from images. _________________________________________________________________ Hertfordshire Computer Science Research Colloquium Abstracts On-line: http://homepages.feis.herts.ac.uk/~nehaniv/colloq/