Title :
Recent advances and challenges in TV structuring
Author_Institution :
Campus Univ. de Beaulieu, INRIA, Rennes, France
Abstract :
TV represents a huge source of data. Even if a TV stream exhibits a strong structure to the viewer, in terms of programs and breaks, this structure is completely implicit in the stream, which is a simple sequence of images and audio frames. This paper presents recent works achieved to recover the structure of such a stream. 4 categories of works are presented, as well as their results and respective requirements in term of annotation. The paper ends by outlining the challenges to be solved in this largely opened field of research.
Keywords :
audio signal processing; image sequences; video streaming; TV structuring; annotation; audio frame; image sequence; structure recovery; Databases; Feature extraction; Hidden Markov models; Manuals; Multimedia communication; Streaming media; TV; TV stream structuring; commercial detection; program detection;
Conference_Titel :
Signal Processing Conference (EUSIPCO), 2012 Proceedings of the 20th European
Print_ISBN :
978-1-4673-1068-0