We present a content-based automatic music tagging algorithm using fully
convolutional neural networks (FCNs). We evaluate different architectures
consisting of 2D convolutional layers and subsampling layers only. In the
experiments, we measure the AUC-ROC scores of the architectures with different
complexities and input types using the MagnaTagATune dataset, where a 4-layer
architecture shows state-of-the-art performance with mel-spectrogram input.
Furthermore, we evaluated the performances of the architectures with varying
the number of layers on a larger dataset (Million Song Dataset), and found that
deeper models outperformed the 4-layer architecture. The experiments show that
mel-spectrogram is an effective time-frequency representation for automatic
tagging and that more complex models benefit from more training data.

本篇论文采用全卷积神经网络实现一种基于内容的自动音乐标记算法，通过在 MagnaTagATune 数据集上的实验，我们发现采用 mel-spectrogram 作为输入，4 层结构的全卷积神经网络在 AUC-ROC 分数方面达到了最先进的性能，但在 Million Song 数据集上则表明，更深的模型优于 4 层结构。实验证明 mel-spectrogram 是一种有效的时频表示方法，更复杂的模型能够从更多的训练数据中受益。