Explore chapters and articles related to this topic
Block-based pseudo-relevance feedback for image retrieval
Published in Journal of Experimental & Theoretical Artificial Intelligence, 2022
Wei-Chao Lin
Relevance feedback can be defined as follows. Suppose n images are stored in a CBIR database for which a query-by-example interface is provided for users to submit a query image for purposes of retrieval. Let Q be the query image example and I a database image, which are represented by feature vectors and , respectively, where m is the number of low-level features extracted. To estimate the visual similarity between Q and I, the normalised Euclidean metric is generally used: