Four Frames Suffice. A Provisionary Model of Vision and Space

摘要

This paper presents a general computational treatment of how mammals are able to deal with visual objects and environments. Among the issues addressed are: constancies and the stable visual world, indexing and context effects, perceptual generalization and allocentric spatial maps. The computational model is expressed in connectionist terms, allowing biological as well as psychological experiments to be included. The model relies heavily on contemporary work in Artificial Intelligence, but is claimed to be consistent with all relevant findings. (Author)