Towards Effective Relevance Feedback Methods in Content Based Image Retrieval Systems

Information Retrieval is one of the major research areas in the recent years. There are two kinds of IR i.e., Content Based and Text Based Retrieval. Text-Based Retrieval is focused on document Retrieval and Content-Based Retrieval is focused on the visual features. It includes audio, video, images, text. Content-Based Information Retrieval includes the CBIR (content-based image retrieval), CBVR (content-based video retrieval) and so on. CBIR system has become a very active research topic during the last few years. To improve the retrieval (text, image, etc.,) performance in content-based image retrieval system, an approach was introduced, named “Relevance Feedback”. It accepts the feedback from the user to retrieve the content which is closest to what he is thinking about. In this paper we discuss the current state-of-the-art in Relevance Feedback as seen from content-based image retrieval point of view and recommend a novel approach for the future.