The SenseCam is a small, wearable device produced by Microsoft Research which is designed to capture thousands of photographs in a single day and which in some way captures or logs the wearer's activities. It also has several on-board sensors which guide when to take photographs, and the sensed information, as well as the photos, are stored on-board for subsequent downloading. A SenseCam is one of a suite of devices recently available which try to capture a wearer's lifelog, or electronic diary of the wearer's activities, and it provides input for a personal digital memory. Developers of technology like the SenseCam have discovered that in creating these new technologies they have also created new problems of how to manage and usefully use the massive amounts of multimedia information created by these devices, or put another way, how to derive semantics from raw multimedia information. Work on deriving meaning from such information represents a subtle shift away from deriving semantics from low-level information to deriving semantics from high-level multimedia information. In this talk we present a set of project activities and results we have ongoing within our research group on gathering semantic information from raw SenseCam and other sensed information. This work contributes to the broader task of managing personal digital memories.
Alan Smeaton is a Professor of Computing and Founding Director of the Centre for Digital Video Processing at Dublin City University. He holds the B.Sc., M.Sc. and PhD degrees in Computer Science from the National University of Ireland. He is a Principal Investigator in the Science Foundation Ireland funded Adaptive Information Cluster, a cross-University grouping of more than 100 researchers working in the broad area of harvesting and using information from diverse sources. Alan has published more than 200 journal, book chapter and conference papers, mostly in the area of information retrieval from diverse media sources. He is strongly associated with the TRECVid evaluation benchmarking campaign, funded by ARDA and run by NIST, which he has been coordinating since it started in 2001. His current research activities are funded by SFI, EU FP6 projects, Enterprise Ireland, Microsoft Research, Google and by other industry partners.
Professor of Intelligent Information Systems
Free University Amsterdam
Guus Screiber studied medicine at the University of Utrecht. After working two years at the University of Leiden in the Medical Informatics department he joined in 1986 the SWI group at the University of Amsterdam, where he was involved in research on knowledge engineering. In 1992 he was awarded a Ph.D. on a thesis entitled "Pragmatics of the Knowledge Level". He has been involved in numerous European and Dutch research projects, including KADS & KADS-II (both on methodologies for knowledge-system development), REFLECT (reflective reasoning), GAMES (medical knowledge systems), KACTUS (technical ontologies), IBROW (Intelligent Brokering on the Web) and MIA (Multimedia Information Analysis). He has published some 100 articles and books. In 2000 he published with MIT Press a textbook on knowledge engineering and knowledge management, based on the CommonKADS methodology. Guus Schreiber is now a professor of Intelligent Information Systems at the Free University Amsterdam. He is chair of the W3C Semantic Web Best Practices and Deployment Working Group and co-chair of W3C's Web Ontology Working Group and member of the Semantic Web Coordination Group of W3C. He is also Scientific Director of the IST Network of Excellence "Knowledge Web".