Project news & updates

Man-Machine Interaction to Facilitate Video Journalism: Researching CASAM

20 May 2008 | PRESS RELEASE


Reporters usually try to avoid pushing the buttons of their sources. But the European Journalism Centre is putting media professionals in a position to do just that. 

The EJC is involved in a multi-national research venture called CASAM, or Computer-Aided Semantic Annotation of Media. The three-year project is an attempt to facilitate and speed up the task of annotating – or “tagging” – video footage for purposes of archiving and retrieval. 

The EJC’s role in this rather scientific project is, of course, the contribution of a journalistic perspective. Our main task will be testing the practical usability of the applications coming out of the engineer’s labs. Media professionals such as librarians, editors and TV journalists will try out the different versions of the system and provide feedback to the developers. 

Automatic image recognition technology is at present still in development stages, not expected to come to full fruition for the next decade. Software is still insufficient when it comes to the identification of context (i.e. semantics). Computers will now recognise well-known individuals, even buildings, and they are able to read typed text and even understand spoken language. However, technology so far can not detect whether for example two politicians in Brussels are struggling or agreeing, or if what they are discussing belongs to one policy area or another. 

CASAM is therefore supposed to develop an intermediate technology bridging the gap between the still dissatisfactory current status and the full-fledged automatism of the future. It will combine human and machine intelligence in such a way that the machines learn how draw on multiple sources of supplementary information in order then use human work capacities most efficiently. 

CASAM is funded by the European Union’s Seventh Research Framework Programme (FP 7) and led by the ICT company, Intrasoft International. Other partners include the universities of Birmingham and Hamburg, the Athens Technology Centre, German international broadcaster Deutsche Welle, Portuguese news agency LUSA and the Greek National Centre of Scientific Research.

CASAM: Computer-Aided Semantic Annotation of Media
The aim of this research and development project was to create a computerised system capable of annotating the vast quantities of video footage with little human intervention, thus transforming the broadcast archive into a valuable sound and image library and saving it from oblivion. EJC’s role in the project was to survey user requirements and perform user tests with CASAM's various prototypes.