Content-based image retrieval systems (panel)

Image Retrieval (IR) problem is concerned with retrieving images that are relevant to users’ requests from a large collection of images, referred to as the image database. A software system that facilitates image retrieval is referred to as the Image Retrieval System (IRS). Previous approaches to the IR problem have been in one of the two directions. In the first direction, image contents are modeled as image attributes. Attributes are extracted manually from the images and are managed within the framework of conventional database systems. The second approach emphasizes the importance of an object recognition system as an integral part of the IRS to overcome the limitations of attribute based retrieval. However, object recognition is a computationally expensive task and renders the approach unsuitable for even the moderate size image databases. Furthermore, IR systems based on this approach tend to be domain-specific. The recent emergence of ubiquitous interest in Multimedia Information Systems has brought the IR problem to the attention of many researchers across several disciplines. Much of this recent research focuses on bridging the gap between the previous two approaches to the problem. The primary emphasis has been on developing domain-independent IRS that provide the ability to retrieve images based on their contents without the need for performing the object recognition task at the query processing time. These efforts have culminated in the introduction of novel image representations and image data models, query processing algorithms for content-based image retrieval, intelligent query languages/interfaces, tools for image database design, and domainindependent system architectures for the IRS. IR Systems that make use of these recent advances are just beginning to appear. Therefore, the theme of the proposed panel is to set up a forum to bring together researchers, developers, and practitioners to explore and discuss various issues concerning the evolving theory, design and development of Content-based Image Retrieval Systems.