A short-term learning approach based on similarity refinement in content-based image retrieval

This paper presents a new relevance feedback approach based on similarity refinement. In the proposed approach weight correction of feature’s components is done by a proposed rule set using mean and standard deviation of feature vectors of relevant (positive) and irrelevant (negative) images. Also, the weight of each type of features is adjusted according to the relevant images’ rank in the retrieval based on only the same type of feature. To evaluate the performance of the proposed method, a set of comparative experiments on a general database containing 20,000 images of various semantic groups are performed. The results confirm the effectiveness of the proposed method comparing with two well-known methods.