Self-supervised learning method using multiple sampling strategies for general-purpose audio representation

TL;DR

This paper introduces a self-supervised audio representation learning method that employs multiple sampling strategies, including novel frame-level and task-specific approaches, leading to improved performance across various audio tasks.

Contribution

The paper presents a new self-supervised learning framework using multiple sampling strategies, enhancing audio representations for diverse tasks beyond conventional clip-level methods.

Findings

Improved clip classification accuracy by 25%.

Enhanced sound event detection by 20%.

Increased pitch detection performance by 3.6%.

Abstract

We propose a self-supervised learning method using multiple sampling strategies to obtain general-purpose audio representation. Multiple sampling strategies are used in the proposed method to construct contrastive losses from different perspectives and learn representations based on them. In this study, in addition to the widely used clip-level sampling strategy, we introduce two new strategies, a frame-level strategy and a task-specific strategy. The proposed multiple strategies improve the performance of frame-level classification and other tasks like pitch detection, which are not the focus of the conventional single clip-level sampling strategy. We pre-trained the method on a subset of Audioset and applied it to a downstream task with frozen weights. The proposed method improved clip classification, sound event detection, and pitch detection performance by 25%, 20%, and 3.6%.