Multi-scale Embedded CNN for Music Tagging (MsE-CNN)

TL;DR

This paper introduces a multi-scale embedded CNN architecture for music tagging that enhances feature transfer across layers, leading to improved classification performance.

Contribution

It proposes a novel CNN model with intermediate connections for better multi-scale feature transfer in music tagging tasks.

Findings

Significant performance improvement over existing methods

Effective transfer of low-level features to final layers

Enhanced multi-scale feature integration

Abstract

Convolutional neural networks (CNN) recently gained notable attraction in a variety of machine learning tasks: including music classification and style tagging. In this work, we propose implementing intermediate connections to the CNN architecture to facilitate the transfer of multi-scale/level knowledge between different layers. Our novel model for music tagging shows significant improvement in comparison to the proposed approaches in the literature, due to its ability to carry low-level timbral features to the last layer.