MATT: A Multiple-instance Attention Mechanism for Long-tail Music Genre Classification. (arXiv:2209.04109v1 [cs.SD])

Imbalanced music genre classification is a crucial task in the Music
Information Retrieval (MIR) field for identifying the long-tail, data-poor
genre based on the related music audio segments, which is very prevalent in
real-world scenarios. Most of the existing models are designed for
class-balanced music datasets, resulting in poor performance in accuracy and
generalization when identifying the music genres at the tail of the
distribution. Inspired by the success of introducing Multi-instance Learning
(MIL) in various classification tasks, we propose …

arxiv attention classification music