direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Page Content



Gefördert durch

EIT ICT Labs activity
01/2011 - 12/2011


OpenSEM, the open innovation platform for se-mantic media, is an EIT ICT Labs activity, started in 2011 with institutes from Germany, Finland, the Netherlands and France.  OpenSEM builds on results achieved in the PetaMedia network of excellence and in some activities closely cooperates with PetaMedia. The consortium is open for new partners and various forms of cooperation  with academic and commercial labs.

OpenSEM helps participants translating research results to product and service innovation. OpenSEM is not just a project: It envisions long term cooperation of institutes and companies working on the cutting edge of semantic multimedia technology, thus establishing a virtual center of excellence on social media retrieval.

Participating Organizations

TU Berlin is represented by the Communication Systems Group and the DAI Labor.

The Communication Systems Group (in German: Fachgebiet Nachrichtenuebertragung) at TUB is directed by Prof. Sikora and comprises of more than 25 researchers. Its teaching and research activities encompass most areas of digital signal processing for multimedia applications. The main research focus is on the content-based analysis, description, segmentation, classification as well as coding of speech, audio, and video data. Prof. Sikora and his group are world-renowned for their contributions to multimedia signal processing and analysis, with many contributions to MPEG-4 and MPEG-7 and a host of peer-reviewed scientific publications to their name.

The DAI-Labor and the chair "Agent technologies in business applications and telecommunication" at Berlin Institute of Technology, headed by Prof. Dr. Sahin Albayrak, perform research and development in order to provide solutions for a new generation of systems and services – "smart services and smart systems". The DAI-Labor applies and tests these solutions in real environments and make them tangible for users. In the context of OpenSEM, two Competence Centers of the DAI-Labor are active.

The Competence Center for Information Retrieval and Machine Learning (CC IRML) of DAI-Labor works on the intelligent acquisition, processing and visualization of data, including frameworks for the semi-automatic collection, extraction and indexing of data, and the development of advanced machine learning algorithms, together with intelligent coordination and cooperation strategies for increased efficiency. The competence center Network & Mobility (CC NEMO) of DAI-Labor also develops and maintains IMS based IPTV and quadruple play solutions, providing a use case for DAI-Labor’s semantic recommendation engine. The teaching and research activities encompass most areas of semantic networking, semantic search, and recommender systems.


TU Delft

Delft University of Technology (TUD) is represented by the Information and Communication Theory (ICT) Group, which has internationally recognized expertise in the area of multimedia content analysis. The research track record of the ICT group in these fields has been built over more than ten years and has enabled continuous and fruitful cooperation with companies such as Philips, British Telecom and Microsoft. In relation to PetaMedia, the ICT Group participated in several EU and Dutch national projects and activities.


VTT Technical Research Centre of Finland is the biggest multitechnological applied research organisation in Northern Europe. VTT provides high-end technology solutions and innovation services.  VTT’s mission is to produce research and innovation services that enhance the international competitiveness of companies, society and other customers. VTT creatively combines its multidisciplinary expertise with the know-how of its partners. VTT is represented by its Media Technologies Knowledge Centre, which has strong background in various media related research topics. The research topics include social and semantic media, user experience, audio & video analysis, video coding, mobile and p2p television, linking print to mobile Internet (=hybrid media), augmented reality, content engineering and visualisation, media production and business models.


The INRIA Rennes - Bretagne Atlantique research centre, organized around 31 research teams, covers all areas of computer science. INRIA's strategic plan guides the research of the Rennes – Bretagne Atlantique centre towards three priority topics:

  • Control of large-scale distributed systems and networks
  • Design, analysis and compilation of embedded software
  • Multimodal images and data: from methodologies to uses

These topics respond to major challenges in the digital society: the Internet of the future in its service dimensions, content and uses; computational sciences serving large sectors such as health, environment and transport. The multimedia team, Texmex, focuses on analyzing and managing collections of multimedia documents with the three following characteristics: large scale collections, multimodal data and semantic analysis. The professional media world (TV, radio, newspapers), with its many diffusion channels, is at the core of Texmex's interests. In this context, the first goal is to develop methods and tools to analyze, structure, describe, and "understand" these documents, in order to get an access to their contents. In particular, research activities are carried out in the fields of image and video copy detection, spoken document processing and multimodal statistical modeling. A second goal is to develop ways to use these documents within application contexts such as information retrieval, navigation in large collections, summarization, repurposing. Integrating security policies is becoming an emerging concern in the team; our long-term goal is to facilitate the use and re-use of the documents in a secure environment.

Summer School on Social Media Retrieval


Multimedia content has become ubiquitous on the web, creating new challenges for indexing, access, search and retrieval. At the same time, much of this content is made available on content sharing websites such as YouTube or Flickr, or shared on social networks like Facebook. In such environments, the content is usually accompanied with metadata, tags, ratings, comments, information about the uploader and their social network, etc. Analysis of these "social media" shows a great potential in improving the performance of traditional multimedia information analysis/retrieval approaches by bridging the semantic gap between the "objective" multimedia content analysis and "subjective" users' needs and impressions. The integration of these aspects however is non-trivial and has created a vibrant, interdisciplinary field of research.   Based on the great success of its previous edition in Interlaken, Switzerland, in 2010 (http://videolectures.net/s3mr2010_interlaken/), The 2nd Summer School on Social Media Retrieval (S3MR) aims at bringing together young researchers from neighbouring disciplines, offering

  1. lectures delivered by experts from academy and industry providing a clear and in-depth summary of state-of-the-art research in social media retrieval, and
  2. collaborative projects in small groups providing hands-on experience on integrative work on selected problems from the field.
  • Content distribution over social/peer-to-peer networks
  • Multimedia content analysis
  • Automatic multimedia annotation/tagging
  • Multimedia indexing/search/retrieval
  • Implicit media tagging
  • Social data analysis
  • Collaborative tagging

Zusatzinformationen / Extras