Metadata and Multilinguality in Video Classification
The VideoCLEF 2008 Vid2RSS task involves the assignment of thematic category labels to dual language (Dutch/English) television episode videos. The University of Amsterdam chose to focus on exploiting archival metadata and speech transcripts generated by both Dutch and English speech recognizers. A Support Vector Machine (SVM) classifier was trained on training data collected from Wikipedia. The results provide evidence that combining archival metadata with speech transcripts can improve classification performance, but that adding speech transcripts in an additional language does not yield performance gains.
KeywordsVideo classification SVM speech recognition
Unable to display preview. Download preview PDF.
- 1.Larson, M., Newman, E., Jones, G.: CLEF 2008 working notes. In: Borri, F., Nardi, A., Peters, C. (eds.) Overview of VideoCLEF 2008: Automatic generation of topic-based feeds for dual language audio-visual content (2008)Google Scholar
- 3.Drucker, H., Wu, D., Vapnik, V.N.: Support vector machines for spam categorization. IEEE Transactions on Neural Networks (5), 1048–1054 (1999)Google Scholar
- 5.Leopold, E., Kindermann, J., Paass, G., Volmer, S., Cavet, R., Larson, M., Eickeler, S., Kastner, T.: Integrated classification of audio, video and speech using partitions of low-level features. In: Proceedings of the Workshop on Multimedia Discovery and Mining (2003)Google Scholar