Overview of MediaEval 2011 Rich Speech Retrieval Task and Genre Tagging Task

The MediaEval 2011 Rich Speech Retrieval Tasks andGenre Tagging Tasks are two new tasks o ered in MediaEval2011 that are designed to explore the development of tech-niques for semi-professional user generated content (SPUG).They both use the same data set: the MediaEval 2010 WildWild Web Tagging Task (ME10WWW). The ME10WWWdata set contains Creative Commons licensed video collectedfrom blip.tv in 2009. It was created by the PetaMedia Net-work of Excellence (http://www.petamedia.eu) in order totest retrieval algorithms for video content as it occurs ‘in thewild’ on the Internet and, in particular, for user contributedmultimedia that is embedded within a social network. Thedata set was initially described in [3]. In this overview pa-per, we repeat the essential characteristics of the data set,describe the tasks and specify how they are evaluated.