Random Projections of Mel-Spectrograms as Low-Level Features for Automatic Music Genre Classification
This provides a computationally efficient alternative for music genre classification without requiring extensive domain expertise, though it is incremental in feature engineering.
The paper tackled music genre classification by using random projections of Mel-spectrograms as low-level features, achieving results comparable to learned features and outperforming transfer learning in shallow learning scenarios across five public datasets.
In this work, we analyse the random projections of Mel-spectrograms as low-level features for music genre classification. This approach was compared to handcrafted features, features learned using an auto-encoder and features obtained from a transfer learning setting. Tests in five different well-known, publicly available datasets show that random projections leads to results comparable to learned features and outperforms features obtained via transfer learning in a shallow learning scenario. Random projections do not require using extensive specialist knowledge and, simultaneously, requires less computational power for training than other projection-based low-level features. Therefore, they can be are a viable choice for usage in shallow learning content-based music genre classification.