MULTIMODAL SEGMENTAL-BASED MODELING OF TENNIS VIDEO BROADCASTS (WedPmPO1)

Author(s) : Manolis Delakis (IRISA/University of Rennes 1, France) Guillaume Gravier (IRISA/CNRS, France) Patrick Gros (IRISA/CNRS, France)

Abstract : Efficient multimodal fusion is a key feature of future video indexing systems. Hidden Markov Models provide a powerful framework for video structure analysis but they require all modalities to be strictly synchronous. Taking as a case study tennis broadcasts analysis, we introduce into video indexing Segment Models, a generalization of Hidden Markov Models, where the fusion of different modalities can be performed in a more flexible way. Segment Models were experimentally proved to perform marginally better compared to Hidden Markov Models.

Menu