Audio Barlow Twins: Self-Supervised Audio Representation Learning
This work addresses audio representation learning for researchers and practitioners, but it is incremental as it applies an existing method to a new domain.
The authors tackled the problem of self-supervised audio representation learning by adapting the Barlow Twins method from computer vision to audio, achieving results that outperform or match the state-of-the-art on 18 tasks from the HEAR 2021 Challenge.
The Barlow Twins self-supervised learning objective requires neither negative samples or asymmetric learning updates, achieving results on a par with the current state-of-the-art within Computer Vision. As such, we present Audio Barlow Twins, a novel self-supervised audio representation learning approach, adapting Barlow Twins to the audio domain. We pre-train on the large-scale audio dataset AudioSet, and evaluate the quality of the learnt representations on 18 tasks from the HEAR 2021 Challenge, achieving results which outperform, or otherwise are on a par with, the current state-of-the-art for instance discrimination self-supervised learning approaches to audio representation learning. Code at https://github.com/jonahanton/SSL_audio.