A collaborative, long-term learning appro ach to using relevance feedback in content-based image retrieval systems [appro ach read approach]

In recent years, an extensive research in the area of content-based image retrieval (CBIR) has been focused on relevance feedback (RF) techniques to improve the retrieval of images. In relevance feedback systems, a search engine dynamically updates the weights of various visual features in the query based on the user's measure of retrieved images' (ir)relevance. In this paper, we propose a novel collaborative relevance feedback approach that relies on the postulation that human-perception subjectivity is narrower than the semantic gap. As relevant results of each query are being remembered by the system, various users over different sessions can contribute to improve the successive query results