Explaining Deep Convolutional Neural Networks On Music Classification

Abstract

Deep convolutional neural networks (CNNs) have been actively adopted in the field of music information retrieval, e.g. genre classification, mood detection, and chord recognition. However, the process of learning and prediction is little understood, particularly when it is applied to spectrograms. We introduce auralisation of a CNN to understand its underlying mechanism, which is based on a deconvolution procedure introduced in [2]. Auralisation of a CNN is converting the learned convolutional features that are obtained from deconvolution into audio signals. In the experiments and discussions, we explain trained features of a 5-layer CNN based on the deconvolved spectrograms and auralised signals. The pairwise correlations per layers with varying different musical attributes are also investigated to understand the evolution of the learnt features. It is shown that in the deep layers, the features are learnt to capture textures, the patterns of continuous distributions, rather than shap

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keychoi2016explaining

Related papers

• Spectral And Rhythm Features For Audio Classification With Deep Convolutional Neural Networks (2024)0.00
• Convolutional Recurrent Neural Networks For Music Classification (2016)18.98
• Multi-level And Multi-scale Feature Aggregation Using Pre-trained Convolutional Neural Networks For Music Auto-tagging (2017)15.43
• Sample-level CNN Architectures For Music Auto-tagging Using Raw Waveforms (2017)13.23
• Music Artist Classification With Convolutional Recurrent Neural Networks (2019)11.93
• Audio-based Music Classification With Densenet And Data Augmentation (2019)10.48
• Sample-level Deep Convolutional Neural Networks For Music Auto-tagging Using Raw Waveforms (2017)0.00
• A Novel Multimodal Music Genre Classifier Using Hierarchical Attention And Convolutional Neural Network (2020)0.00