DMAG-Search and Retrieval (DMAG-Search and Retrieval)


Project description

Multimedia Information Retrieval (MIR) is a branch of Information Retrieval where the information is stored not only in the form of text, but also in the form of any other digital content such as still and moving images or audio. MIR is the science of searching for digital contents, for information within digital contents, and for metadata about digital contents. MIR is interdisciplinary, and is related with structured data retrieval (e.g. databases), information retrieval from unstructured data (e.g. document retrieval), data integration, pattern recognition, cluster analysis, image and audio analysis, and knowledge representation and discovery. Multimedia Information Retrieval is a critical area of study today, the search for digital contents is one of the major challenges of our time. Digital libraries, healthcare, bio-computing, social networking sites or cultural heritage are just some examples which have created a worldwide need for new paradigms and techniques on how to browse, search and summarize multimedia collections and more generally, how to afford efficient multimedia content consumption.

DMAG’s Projects

DMAG’s work in this area has been traditionally focused in algorithms and data structures for multimedia metadata representation and metadata-based multimedia retrieval. Evolving from XML to Semantic Web technologies, the group began focusing in the distributed retrieval problem, applied to federating multiple commercial vertical web search engines such as online news portals or academic publishing databases. Lately, the group has been involved in the design and standardization of formats and protocols to facilitate the interoperable access to multimedia repositories. DMAG is playing a key role in the standardization of the MPEG Query Format (ISO/IEC 15938-12:2008) and JPSearch (ISO/IEC 24800). While the MPEG Query Format is a universal language for querying multimedia databases, JPSearch is a new norm from JPEG which aims standardizing the different aspects related to the interchange, management and search of still images and still images repositories. One of the last active projects of the group related to these standards consists in developing a medical image search engine for querying optical biopsies obtained through confocal endomicroscopy. The resulting system will aid diagnosing digestive diseases minimizing the need of surgery, and thus minimizing cost and improving patients’ experience.