Self-supervised learning for Environmental Sound Classification

Achyut Mani Tripathi

doi:10.1016/j.apacoust.2021.108183

Return

Self-supervised learning for Environmental Sound Classification

2021-11-01

PRE

Achyut Mani Tripathi *

Aakansha Mishra

DOI：10.1016/j.apacoust.2021.108183

Original

Save

Abstract

En 中文

Environmental Sound Classification (ESC) is one of the most challenging tasks in signal processing, digital forensic and machine learning. Numerous methods have been proposed to perform ESC. The conventional models' training depends on an enormous amount of annotated data, specifically while training the deep models. This paper presents a self-supervised learning (SSL)-based deep classifier for ESC, which is an under-explored method in the field of ESC. SSL mechanism directs the model to effectively learn prototypical features from the data itself by solving a pretext task. The model proposed in this paper takes spectrogram images as input. A pretext or an auxiliary task is defined as identification of the type of data augmentation applied to the signal. The model learned by solving the pretext task is further fine-tuned for developing the deep model for ESC. The model's performance is evaluated on two benchmark sound classification datasets, i.e. ESC-10 and DCASE 2019 Task-1(A) datasets. The experiments and results show that the SSL model attains an improvement of 12.59% and 11.17% in accuracy compared to the baseline models of the DCASE 2019 Task-1(A) and ESC-10 datasets respectively. Moreover, the model also shows competitive performance to state-of-the-art methods. (C) 2021 Elsevier Ltd. All rights reserved.

Keywords:

Environmental Sound Classification

Data Augmentation

Residual Network

Self-Supervised Learning

Sound

Transfer Learning