A computational model of color perception and color naming

I define a computational model of color perception and color naming which is based in part on neurophysiological data and which can explain results in psychological, anthropological, and linguistic studies on color naming and categorization. In particular, the model explains the graded nature and foci of color categories. This model constitutes a semantic model of basic color terms, grounded in perception. It allows an artificial cognitive agent to name color samples, point out examples of named colors in its environment, and select objects from its environment by color, consistent with human performance on the same tasks. An algorithm and implementation for the computational model is presented, and evaluated theoretically and experimentally. The contributions of this work are in autonomous agent architecture on the one hand, particularly in the area of symbol grounding and embodiment, and in perceptual modeling, particularly color vision, on the other hand. The application presented is a vertically integrated one, ranging from real visual input to symbolic description, and back.