- From: Hausenblas, Michael <michael.hausenblas@joanneum.at>
- Date: Fri, 5 Jan 2007 21:15:51 +0100
- To: Raphaël Troncy <Raphael.Troncy@cwi.nl>
- Cc: <public-xg-mmsem@w3.org>
Raphaël, All, Happy New Year to all of you as well. Thanks for this pointer--looks great; I'm _very_ eager to see it working :) though I wonder what they use to do the face/object recognition and the so called "intelligent automatic tagging" ... In this context I was contemplating about (automated) description of real-world media; imagine not only still images, but video clips being "tagged intelligent"--what could the requirements for the KR-part look like? The following example might give you an idea what causes me some headache: Take a video clip with a duration of one hour that is described with MPEG-7. Several visual features (F) as colour, shape, texture, etc. are extracted for a number of spatial segments (S) per key frame (K). Further, a kind of multimedia ontology (e.g. [1]) is then used to represent the MPEG-7 descriptors formally (viz. on a RDF/OWL basis); an average number of RDF triples is assumed for each descriptor (TD). An estimation of the resulting RDF graph size then is F*K*S*TD. Let us assume that we want to capture 10 features, some 1000 key frames may exist, 10 spatial segments are marked up, and finally 10 triples are required per descriptor. This yields a total RDF graph size of _1 million triples_, just for describing some low-level features of an hour of video footage. What do you think? Cheers, Michael [1] http://www.w3.org/2005/Incubator/mmsem/wiki/Vocabularies#f_MPEG-7 ---------------------------------------------------------- Michael Hausenblas, MSc. Institute of Information Systems & Information Management JOANNEUM RESEARCH Forschungsgesellschaft mbH Steyrergasse 17, A-8010 Graz, AUSTRIA <office> phone: +43-316-876-1193 (fax:-1191) e-mail: michael.hausenblas@joanneum.at web: http://www.joanneum.at/iis/ <https://webmail.joanneum.at/exchweb/bin/redir.asp?URL=http://www.joanneum.at/iis/> <private> mobile: +43-660-7621761 web: http://www.sw-app.org/ <https://webmail.joanneum.at/exchweb/bin/redir.asp?URL=http://www.sw-app.org/> ---------------------------------------------------------- ________________________________ From: public-xg-mmsem-request@w3.org on behalf of Raphaël Troncy Sent: Thu 2007-01-04 20:19 To: MMSem-XG Public List Cc: Susanne Boll Subject: Ookles Dear MMSemers, First of all, let me wish to all of you an Happy New Year 2007 ! Something interesting for the Multimedia Semantics XG, the Ookles platform should be launched in 2007, see: http://www.ookles.com/ Ookles can be defined by the following equation: Ookles = Flickr+Riya+YouTube An interesting technical review can be read at: http://www.techcrunch.com/2006/12/18/ookles-to-launch-in-early-2007/ Among others, Ookles propose to organize automatically your photos and has developped the following features: * Facial and object recognition finds the people you care about * Automatic organization of your photos * Intelligent automatic tagging for your media * Controls for security and viewer safety * Import and export from all popular photo and video sharing websites * All in an easy web service accessible anywhere Best regards. Raphaël -- Raphaël Troncy CWI (Centre for Mathematics and Computer Science), Kruislaan 413, 1098 SJ Amsterdam, The Netherlands e-mail: raphael.troncy@cwi.nl & raphael.troncy@gmail.com Tel: +31 (0)20 - 592 4093 Fax: +31 (0)20 - 592 4312 Web: http://www.cwi.nl/~troncy/
Received on Friday, 5 January 2007 20:16:31 UTC