<XML><RECORDS><RECORD><REFERENCE_TYPE>3</REFERENCE_TYPE><REFNUM>8692</REFNUM><AUTHORS><AUTHOR>Ren,R.</AUTHOR><AUTHOR>Punitha,P.</AUTHOR><AUTHOR>Jose,J.M.</AUTHOR></AUTHORS><YEAR>2007</YEAR><TITLE>Attention based video summarisation in Rushes Collection</TITLE><PLACE_PUBLISHED>ACM Multimedia Information Retrieval (MIR 07)</PLACE_PUBLISHED><PUBLISHER>N/A</PUBLISHER><LABEL>Ren:2007:8692</LABEL><KEYWORDS><KEYWORD>Multimedia retrieval</KEYWORD></KEYWORDS<ABSTRACT>This paper presents the framework of a general video summarisation system on the rushes collection, which formalises the summarisation process as an 0-1 knapsack optimisation problem. Three stages are included, namely content analysis, content selection and composition. Content analysis is the pre-processing step, consisting of shot segmentation, feature extraction, raw video discrimination and shot clustering. Content selection weoghts the importance of video segments by an attention model. A greedy approximation approach is employed in the composition of summary video with the cost function, which balanaces the video importance gain and the duration cost. Te average content coverage achieved on the rush test collection is about 29%, while the average qulification score on readability is 3.13 with the redundancy credit at 4.08.</ABSTRACT></RECORD></RECORDS></XML>