TY - GEN
T1 - LEARNING MULTI-LEVEL REPRESENTATIONS FOR HIERARCHICAL MUSIC STRUCTURE ANALYSIS
AU - Buisson, Morgan
AU - McFee, Brian
AU - Essid, Slim
AU - Crayencour, Hélène C.
N1 - Publisher Copyright:
© M. Buisson, B. McFee, S. Essid and H. C. Crayencour.
PY - 2022/1/1
Y1 - 2022/1/1
N2 - Recent work in music structure analysis has shown the potential of deep features to highlight the underlying structure of music audio signals. Despite promising results achieved by such representations, dealing with the inherent hierarchical aspect of music structure remains a challenging problem. Because different levels of segmentation can be considered as equally valid, specifically designed representations should be optimized to improve hierarchical structure analysis. In this work, unsupervised learning of such representations using a contrastive approach operating at different time-scales is explored. The proposed system is evaluated on flat and multi-level music segmentation. By leveraging both time and the hierarchical organization of music structure, we show that the obtained deep embeddings can encode meaningful patterns and improve segmentation at various levels of granularity.
AB - Recent work in music structure analysis has shown the potential of deep features to highlight the underlying structure of music audio signals. Despite promising results achieved by such representations, dealing with the inherent hierarchical aspect of music structure remains a challenging problem. Because different levels of segmentation can be considered as equally valid, specifically designed representations should be optimized to improve hierarchical structure analysis. In this work, unsupervised learning of such representations using a contrastive approach operating at different time-scales is explored. The proposed system is evaluated on flat and multi-level music segmentation. By leveraging both time and the hierarchical organization of music structure, we show that the obtained deep embeddings can encode meaningful patterns and improve segmentation at various levels of granularity.
M3 - Conference contribution
AN - SCOPUS:85164948321
T3 - Proceedings of the 23rd International Society for Music Information Retrieval Conference, ISMIR 2022
SP - 591
EP - 597
BT - Proceedings of the 23rd International Society for Music Information Retrieval Conference, ISMIR 2022
A2 - Rao, Preeti
A2 - Murthy, Hema
A2 - Srinivasamurthy, Ajay
A2 - Bittner, Rachel
A2 - Repetto, Rafael Caro
A2 - Goto, Masataka
A2 - Serra, Xavier
A2 - Miron, Marius
PB - International Society for Music Information Retrieval
T2 - 23rd International Society for Music Information Retrieval Conference, ISMIR 2022
Y2 - 4 December 2022 through 8 December 2022
ER -