Shot classification using domain specific features for movie management

Publication Type:
Conference Proceeding
Citation:
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2012, 7239 LNCS (PART 2), pp. 314 - 318
Issue Date:
2012-05-11
Metrics:
Full metadata record
Files in This Item:
Filename Description Size
Thumbnail2012004147OK.pdfPublished Version401.97 kB
Adobe PDF
Among many video types, movie content indexing and retrieval is a significantly challenging task because of the wide variety of shooting techniques and the broad range of genres. A movie consists of a series of video shots. Managing a movie at shot level provides a feasible way for movie understanding and summarization. Consequently, an effective shot classification is greatly desired for advanced movie management. In this demo, we explore novel domain specific features for effective shot classification. Experimental results show that the proposed method classifies movie shots from wide range of movie genres with improved accuracy compared to existing work. © 2012 Springer-Verlag.
Please use this identifier to cite or link to this item: