Music genre classification is an essential tool for music informationretrieval systems and it has been finding critical applications in variousmedia platforms. Two important problems of the automatic music genreclassification are feature extraction and classifier design. This paperinvestigates inter-genre similarity modelling (IGS) to improve the performanceof automatic music genre classification. Inter-genre similarity information isextracted over the mis-classified feature population. Once the inter-genresimilarity is modelled, elimination of the inter-genre similarity reduces theinter-genre confusion and improves the identification rates. Inter-genresimilarity modelling is further improved with iterative IGS modelling(IIGS) andscore modelling for IGS elimination(SMIGS). Experimental results with promisingclassification improvements are provided.
展开▼
