Learning color-appearance models by means of feed-forward neural networks
Paola Campadelli, Cristina Gangai, Raimondo Schettini · Color Research & Application · 1999
Device-independent color imaging demands a reliable color-appearance model. We present a method for faithfully approximating color-appearance models by means of feed-forward neural networks trained with the error back-propagation algorithm. In particular, we present experimental evidence that in several “standard” viewing conditions recommended for testing color-appearance models, the same network architecture is capable of learning quite satisfactorily the transformations performed by different color-appearance models. © 1999 John Wiley & Sons, Inc. Col Res Appl, 24, 411–421, 1999