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

Jordi Pons, Xavier Serra
9/14/2019
cs.SDcs.CLeess.AS

Official code: jordipons/musicnn (724 stars, ISC, last commit 2023-12-11). 2 more implementations below.

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.

View on arXivView PDF

Code Implementations(3)

jordipons/musicnnOfficial100%
ISC

Pronounced as "musician", musicnn is a set of pre-trained deep convolutional neural networks for music audio tagging.

724103Python, Jupyter NotebookJul 8, 20192 years agoISC
jordipons/musicnn-trainingOfficial100%
ISC

Tensorflow code for training deep convolutional neural networks for music audio tagging

5114C++, PythonJul 19, 20193 years agoISC
jordipons/sklearn-audio-transfer-learningCommunity100%
ISC

A didactic toolkit to rapidly prototype audio classifiers with pre-trained Tensorflow models and Scikit-learn

14929PythonMay 8, 20193 years agoISC

Similar Papers

Convolutional Recurrent Neural Networks for Music Classification

88% match1 repo132016

Sample-level CNN Architectures for Music Auto-tagging Using Raw Waveforms

86% match1 repo582017

Evaluation of CNN-based Automatic Music Tagging Models

86% match1 repo4402020

General audio tagging with ensembling convolutional neural network and statistical features

85% match1 repo782018

Perceptual Musical Features for Interpretable Audio Tagging

85% match1 repo52023

PANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognition

84% match2 repos1.8k2019

Cite this paper

@article{pons2019musicnn,
  title  = {musicnn: Pre-trained convolutional neural networks for music audio tagging},
  author = {Jordi Pons and Xavier Serra},
  year   = {2019},
  eprint = {1909.06654},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1909.06654}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1909.06654.svg)](https://codeofpaper.com/paper/1909.06654)

Discussion