Combining multiple experts for classifying shot changes in video sequences
- 20 January 2003
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2, 700-704
- https://doi.org/10.1109/mmcs.1999.778569
Abstract
The paper presents a shot cut detection technique using a combination of multiple experts. The experts themselves are stand-alone methods to detect shot cuts. We propose a combination method that gives significantly better results compared to these experts on their own. We achieve this by exploiting the complementarity of expert knowledge, i.e. the fact that the various experts calculate different features of the video sequences.Keywords
This publication has 14 references indexed in Scilit:
- Joint spatial-spectral indexing for image retrievalPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Video scene change detection using the generalized sequence tracePublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Boosting the margin: a new explanation for the effectiveness of voting methodsThe Annals of Statistics, 1998
- On combining classifiersPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1998
- Combining classifiers: A theoretical frameworkPattern Analysis and Applications, 1998
- Realisable Classifiers: Improving Operating Performance on Variable Cost ProblemsPublished by British Machine Vision Association and Society for Pattern Recognition ,1998
- A comparative study of decision combination strategies for a novel multiple-expert classifierPublished by Institution of Engineering and Technology (IET) ,1997
- Bagging predictorsMachine Learning, 1996
- Digital video segmentationPublished by Association for Computing Machinery (ACM) ,1994
- Automatic partitioning of full-motion videoMultimedia Systems, 1993