Music theme recognition using CNN and self-attention

2019/11/16 by Manoj Sukhavasi, Sukhavasi, Manoj, Sainath Adapa +1
Computer Science · #Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Handwritten Text Recognition Techniques #Machine Learning (cs.LG) #Music and Audio Processing #Sound (cs.SD) #Speech Recognition and Synthesis #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.1911.07041

openalex publication_date 2019/11/16 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

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.

Citations

• Adam: A Method for Stochastic Optimization
• Automatic tagging using deep convolutional neural networks
• Convolutional Recurrent Neural Networks for Music Classification
• mixup: Beyond Empirical Risk Minimization
• Improving Generalization Performance by Switching from Adam to SGD

Related

• Emotion and Theme Recognition in Music with Frequency-Aware RF-Regularized CNNs
• Visualizing and Understanding Self-attention based Music Tagging
• Sample-level CNN Architectures for Music Auto-tagging Using Raw Waveforms
• Emotion Detection and Music Recommendation System
• Stacked Convolutional and Recurrent Neural Networks for Music Emotion Recognition