Music theme recognition using CNN and self-attention

Abstract

We present an efficient architecture to detect mood/themes in music tracks on autotagging-moodtheme subset of the MTG-Jamendo dataset. Our approach consists of two blocks, a CNN block based on MobileNetV2 architecture and a self-attention block from Transformer architecture to capture long term temporal characteristics. We show that our proposed model produces a significant improvement over the baseline model. Our model (team name: AMLAG) achieves 4th place on PR-AUC-macro Leaderboard in MediaEval 2019: Emotion and Theme Recognition in Music Using Jamendo.

MediaEval 2019, 27-29 October 2019, Sophia Antipolis, France

References in corpus (7)

• Adam: A Method for Stochastic Optimization
• mixup: Beyond Empirical Risk Minimization
• SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition
• MobileNetV2: Inverted Residuals and Linear Bottlenecks
• Improving Generalization Performance by Switching from Adam to SGD
• Automatic tagging using deep convolutional neural networks
• Convolutional Recurrent Neural Networks for Music Classification

Cited by in corpus (1)

• Receptive Field Regularization Techniques for Audio Classification and Tagging with Deep Convolutional Neural Networks