Combining High-Level Features of Raw Audio Waves and Mel-Spectrograms for Audio Tagging

Abstract

In this paper, we describe our contribution to Task 2 of the DCASE 2018 Audio Challenge. While it has become ubiquitous to utilize an ensemble of machine learning methods for classification tasks to obtain better predictive performance, the majority of ensemble methods combine predictions rather than learned features. We propose a single-model method that combines learned high-level features computed from log-scaled mel-spectrograms and raw audio data. These features are learned separately by two Convolutional Neural Networks, one for each input type, and then combined by densely connected layers within a single network. This relatively simple approach along with data augmentation ranks among the best two percent in the Freesound General-Purpose Audio Tagging Challenge on Kaggle.

Detection and Classification of Acoustic Scenes and Events 2018 (DCASE 2018), 19-20 November 2018, Surrey, UK

References in corpus (3)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• mixup: Beyond Empirical Risk Minimization
• General-purpose Tagging of Freesound Audio with AudioSet Labels: Task Description, Dataset, and Baseline

Cited by in corpus (4)

• Do Autonomous Agents Benefit from Hearing?
• nnAudio: An on-the-fly GPU Audio to Spectrogram Conversion Toolbox Using 1D Convolution Neural Networks
• Label-efficient audio classification through multitask learning and self-supervision
• An End-to-End Audio Classification System based on Raw Waveforms and Mix-Training Strategy