Full title

Peer-to-peer Tagged Media

Project website

www.petamedia.eu

Duration of project

42 months, starting from March 2008

Funding source

7th European Frame Programme (EU FP7)

Funding number

EU FP7 contract Nb: 216444

Description

The PETAMEDIA NoE aims at fostering international collaborative research in the areas of multimedia content analysis (MCA) and social and peer-to-peer (SP2P) networks by coordinating the research activities among partner countries, so as to establish a European network composed of national networks, while enabling as well the participation of specific additional interested partners. PETAMEDIA is so composed of four core partners known for their proven track record and coming from four different European states, complemented by four strong national networks and by the interested partners, cf. the figure below. The final purpose of the consortium is to establish a sustainable European virtual centre of excellence connecting thematically interested research groups throughout Europe.

Research activities

The collective research effort will be directed towards integration of existing MCA and SP2P technologies, and towards identification and exploration of potentials and limitations of MCA/SP2P combinations. A particular scientific challenge is the synergetic combination of user-based collaborative tagging, peer-to-peer networks and multimedia content analysis. Solutions and collaborative research field trials will be built on the coordinating partner's open source P2P software Tribler. These field trials are large scale experiments involving real users, who aim at evaluating two aspects, namely:

• New technological contributions to merging automated video content analysis and user-based tagging in a P2P network structure;


• Perception of users of this new form of collaboration in a social network structure build on top of a P2P network structure.

• Development of the technology roadmap and integrated architecture;


• Research on the future of the SP2P/MCA combination;


• Experiments investigating validity and performance of the new SP2P/MCA paradigm;

Social Media Content Acquisition

The main goal of the Integrated Research Project "Social Media Content Acquisition" is designing and collecting a set of information generated by a social network (e.g., Twitter, Digg, Delicious), such as associated implicit and explicit metadata (tags, comments, data, ratings, preferences), associated friendship/trust relationships, etc.

The basic idea is to search social networks for content uploaded from some video share sites, such as Blip.tv or Revver.com which contain video clips under common creative license. In that sense Twitter, Digg, FriendFind or Delicious are crawled for content which comes from already mentioned video share sites. When such content is found, we crawl all data available for the user who uploaded that video, mainly: tags, comments, data, ratings, preferences, etc., but also the social network data for the certain number of levels, e.g. for the friends of that user and friends of friends, etc.

The challenge is to acquire a data set that is not only large enough to allow meaningful research but also contains information generated by a social network of sufficient density. This dataset will be used by PetaMedia community.

Partners

Technical University Delft (TUD)
Queen Mary University London (QMUL)
Technical University of Berlin (TUB)

Partnership structure of PetaMedia

Associated national networks

NL: Netherlands Institute for Research on ICT (NIRICT)
CH: Interactive Multimodal Information Management (IM2)
UK: Multimedia Knowledge Management network (MMKM)
DE: Human-Centric Communication Cluster (HC3)

Selected publications

* P.  Vajda, L. Goldmann, and T. Ebrahimi, "Analysis of the limits of graph-based object duplicate detection," in Proc. Int. Symp. Multimedia, 2009

* A. Yazdani, J.-S. Lee, and T. Ebrahimi, "Implicit emotional tagging of multimedia using EEG signals and brain-computer interface," in Proc. ACM Multimedia, Workshop on Social Media, 2009

* J.-S. Lee, F. De Simone, and T. Ebrahimi, "Influence of audio-visual attention on perceived quality of standard definition multimedia content," in Proc. Workshop on Quality of Multimedia Experience, 2009

* J.-S. Lee, F. De Simone, and T. Ebrahimi, "Video coding based on audio-visual attention," in Proc. Int Conf. Multimedia and Expo, 2009

* P. Vajda, F. Dufaux, T. Ha M., and T. Ebrahimi, "Graph-based approach for 3D object duplicate detection," in Proc. Int. Workshop on Image Analysis for Multimedia Interactive Services, 2009