musicnn: Pre-trained convolutional neural networks for music audio tagging

TL;DR

The paper introduces musicnn, a library of pre-trained convolutional neural networks designed for music audio tagging, offering out-of-the-box tools, feature extraction, and transfer learning capabilities, with state-of-the-art results on benchmark datasets.

Contribution

It presents a set of pre-trained music-specific CNN models and code for training and implementing novel architectures, advancing music audio tagging techniques.

Findings

Achieved 90.77 ROC-AUC on MagnaTagATune

Achieved 88.81 ROC-AUC on Million Song Dataset

Demonstrated effectiveness of attention-based models

Abstract

Pronounced as "musician", the musicnn library contains a set of pre-trained musically motivated convolutional neural networks for music audio tagging: https://github.com/jordipons/musicnn. This repository also includes some pre-trained vgg-like baselines. These models can be used as out-of-the-box music audio taggers, as music feature extractors, or as pre-trained models for transfer learning. We also provide the code to train the aforementioned models: https://github.com/jordipons/musicnn-training. This framework also allows implementing novel models. For example, a musically motivated convolutional neural network with an attention-based output layer (instead of the temporal pooling layer) can achieve state-of-the-art results for music audio tagging: 90.77 ROC-AUC / 38.61 PR-AUC on the MagnaTagATune dataset --- and 88.81 ROC-AUC / 31.51 PR-AUC on the Million Song Dataset.