TUD at MediaEval 2012 genre tagging task: Multi-modality video categorization with one-vs-all classifiers
Peng Xu, Yangyang Shi, Martha A. Larson · MediaEval · 2012
In this paper, we investigate the internet video categorization problems on genre related labels. The videos are represented by features extracted from different modalities. Then for each category, one-vs-all SVM classifiers are trained based on features from different modalities. The weighted Reciprocal Rank Fusion method is used to combine the classifiers for each modality. The experiments are based on the data from the Genre Tagging Task of MediaEval 2012.