Relevance feedback is a power tool to increase the retrieval performance in content-based image retrieval system. In this paper,we review three relevance feedback algorithms,i. e. spatial transform model ,clustering model and Bayes model,and discuss their advantages and disadvantages.
