CumInCAD is a Cumulative Index about publications in Computer Aided Architectural Design
supported by the sibling associations ACADIA, CAADRIA, eCAADe, SIGraDi, ASCAAD and CAAD futures

id b83a
authors Benoudjit A, Derix C and Coates P
year 2004
title Human perception and space classification: The Perceptive Network
source Proceedings of the Generative Arts conference, Milan, 2004
summary This paper presents a computer model for space perception, and space classification that is built around two artificial neural networks (ANN). This model is the first known application in architecture, where a self-organized map (SOM) is used to create a space classification map on the base of human perception criteria. This model is built with the aim to help both the space designers (architects, interior designer and urban designers), and the space users to gain a better understanding of the space in particular, and the environment where they evolve in general. This work is the continuity of an outgoing work started in the CECA by C. Derix around Kohonen network.
keywords neural network, self-organised feature map, perception, spatial configuration
series other
type normal paper
full text file.pdf (916,489 bytes)
references Content-type: text/plain
last changed 2012/09/20 19:28
HOMELOGIN (you are user _anon_148719 from group guest) Works Powered by SciX Open Publishing Services 1.002