European Workshop on the Integration of Knowledge, Semantics and Digital Media Technology (EWIMT 2004) November 25 – 26, London, UK The workshop turned out to be not as low-level feature oriented than expected, as there were relevant papers on semantic multimedia. Met quite a number of people from my MPEG-7 days but also Eric Pauwels from CWI, Guus and Ian Horrocks (could not stay for the panel itself and listened only to the panelists statements, though). Also talked to Jerome Euzenat from INRIA – Rhone-Alpes, the supervisor of Raphael. They are doing a lot of knowledge representation on rhetoric and discourse and it seems that in particular Stefano and Katya should have a look. The outcome of that workshop is: we need high-level semantic descriptions (surprise) and most of the low-level feature researchers realize that too (enough papers on how to bridge the semantic gab – were most of course don't bridge but rather stroll along the gab and steal other people's time by filling 20 – 45 minutes with useless if not dangerously false presentations - see the talk by Tsuhan Chen for the first and the paper by Ren et al. for the latter). Interesting to see – there are more people doing similar things like us despite the one's we already know ..... and ..... we are on the correct track and are still ahead of the pack. What is required, though, that we have to publish more (our work is not as widely known as it could be). Short description of relevant papers/posters follows (proceedings-cd on my desk – note, it was a relatively expensive workshop but printed proceedings only for the presenters – hm). First day: A Framework for Semantic Web Enabled Multimedia Lyndon Nixon (Freie Universität Berlin) The talk was closely related to the work Joost was doing together with Jane and the follow up with Disc. Topic – how to use the semantic web to overcome ambiguity while searching for material that should be used in presentations. He also mentioned Cuypers in his presentation. Might be useful if Joost has a look at what Lyndon is up to. Besides, Lyndon is currently finishing his PhD on that topic too. Lynda, he mentioned that he contacted you some time ago and that you exchanged mails. Perhaps it might be a good idea if he comes over and gives a talk about his work. Problem: no proper answer to the question who is providing the annotations he requires in his framework (not during the question answering time nor later). He could also not clearly specify how his user profile is structured and during a later conversation it became clear that that it is not implemented yet. In the first poster session was one presentation Stefano should have a look at: E-movie Creation by Rule-Based Reasoning from the Director's Viewpoint. The described system automatically generates an animated video from a given textual story where the emphasis is to find the best frame size and camera position. The system makes use of knowledge about mise-en-scene, cinematography, and montage. Though the work is presenting the last stage of a PhD I am a bit in doubt, as the student had no idea about the work on video-based knowledge representation, automatic video generation and automatic camera orientation done in the 90. She was a bit in shock after the list of references she should look at. In the second poster session the work of T. Cleenewerck was of interest. The poster was about the IMEDIA project, emphasising the issues of evolutionary development of quiz shows in interactive TV (automatically generated). The language part might interest Jacco and Joost. In the last poster session of the first day (sorry, the paper sessions were utterly boring\ - and no wireless connection or any other connections to the web – so I read Ed Hartley's PhD thesis instead) the work of M. K. Asadi was very close to what we are doing. The title of the poster: Knowledge-based and Semantic Adaptation Of Multimedia Content. The approach is to annotate media items in an MPEG-21 (altered) and MPEG-7 framework, with focusing on such elements of semantic relations between media items, their role in a discourse, etc. The final output is in SMIL (her presentation even look very much like ours). She is currently writing he PhD (another 7 month to go). It might make sense to invite her for a presentation too. The other potentially interesting poster about Domain Ontology building for Semantic Web Applications to Cultural Heritage Repositories was unfortunately not presented. Second day: The presentations by the EU commission did not provide any new information with respect to how to prepare a call. More interesting are the topics for the current call (deadline 20th of March): - knowledge acquisition and modelling - knowledge sharing and reuse - intelligent content Drawback: if the commission talks here about users they have organisations in mind. Relevant for us: the STREP instrument -> single-problem R&D, high risk, 5-6 partners, 3-4 countries most favoured for research on intelligent content The remaining presentations of the day were neglectable. Interesting poster: NM2 : NEW MEDIA FOR A NEW MILLENNIUM DR. D. L. WILLIAMS AND I. C. KEGEL British Telecommunications plc This paper describes NM2, a project which will create prototype production tools for the media industry that will allow the easy production of non-linear media genres based on moving audiovisual images suitable for transmission over broadband networks. There is, however, quite a gab between the provided outline of the project and their actual understanding of story structures. In any case, it seems that Stefano should contact them. KNOWLEDGE-BASED EXPLORATION OF MULTIMEDIA MUSEUM COLLECTIONS S. GOODALL, P. H. LEWIS, K. MARTINEZ, P. A. S. SINCLAIR, F. School of Electronics and Computer Science, University of Southampton, UK SCULPTEUR is an EU/IST funded collaborative research project concerned with the application of knowledge and semantic web technologies to the problem of multimedia information management in the cultural heritage domain. To some extend there is some overlap here with SampLe. Also had a longer chat with M. Addis, who is manager of the IT Innovation centre, and he showed great interest in Stefano's work (from the editing side).