An in-memory relevance feedback technique for high-performance image retrieval systems

Abstract

Content-based image retrieval with relevant feedback has been widely adopted as the query model of choice for improved effectiveness in image retrieval. The effectiveness of this solution, however, depends on the efficiency of the feedback mechanism. Current methods rely on searching the database, stored on disks, in each round of relevance feedback. This strategy incurs long delay making relevance feedback less friendly to the user, especially for very large databases. Thus, scalability is a limitation of existing solutions. In this paper, we propose an in-memory relevance feedback technique to substantially reduce the delay associated with feedback processing, and therefore improve system usability. Our new data-independent dimensionality-reduction technique is used to compress the metadata to build a small in-memory database to support relevance feedback operations with minimal disk accesses. We compare the performance of this approach with conventional relevance feedback techniques in terms of computation efficiency and retrieval accuracy. The results indicate that the new technique substantially reduces response time for user feedback while maintaining the quality of the retrieval.