Content-based image retrieval systems require the development of relevance
feedback mechanisms that allow the user to progressively refine the system'
s response to a query. In this paper a new relevance feedback mechanism is
described which evaluates the feature distributions of the images judged re
levant, or not relevant, by the user and dynamically updates both the simil
arity measure and the query in order to accurately represent the user's par
ticular information needs. Experimental results demonstrate the effectivene
ss of this mechanism. (C) 1999 Elsevier Science Ltd. All rights reserved.