The research on multimedia retrieval focuses on the development of tools and techniques to support the easy access to information in digital archives. It exploits the synergy that exists between language & speech technology, image processing and database technology. The results can be used in innovative applications that support various content management tasks, e.g. automated metadata extraction and indexing of large text collections and audiovisual archives, search technology and filtering of dynamic information streams.
Applications built on this type of tools and expertise can adapt to specific domains, but the underlying techniques are domain-ndependent and can be integrated in any application platform.
HMI participates in several national and international collaborative projects in the context of which multimedia retrieval issues are investigated, e.g. Druid, Waterland, Echo, MUMIS, M4, MultimediaN, AMI, MESH, MEdiaCampaign, AMIDA. Within these projects HMI works together with both technology providers, as well as content owners, such as broadcast archives. Methods developed are regularly evaluated in international benchmark events such as CLEF, TREC, TRECVID, INEX and TDT.
Focal research themes within HMI are:
- Cross-lingual information retrieval
- Crossmodal indexing
- Information extraction
- Ontology processing
- Spoken document retrieval
- Automatic speech recognition for Dutch, plus a number of EU languages
Students can participate via individual assignments and via stages.
The following HMI-members are working on Multimedia Retrieval:Former HMI-members:
Here you can find the publications
Students can find information about assignments:
[ Capita selecta and Research Topics ] at our Msc webpages.