Bootstrap your own latent: A new approach to self-supervised Learning

TL;DR

BYOL introduces a self-supervised learning method using two networks that learn from each other without negative pairs, achieving state-of-the-art image classification accuracy.

Contribution

It presents a novel self-supervised learning approach that outperforms previous methods relying on negative pairs, using a bootstrap mechanism with online and target networks.

Findings

Achieves 74.3% top-1 accuracy on ImageNet with ResNet-50

Performs on par or better on transfer and semi-supervised benchmarks

Does not require negative pairs for effective learning

Abstract

We introduce Bootstrap Your Own Latent (BYOL), a new approach to self-supervised image representation learning. BYOL relies on two neural networks, referred to as online and target networks, that interact and learn from each other. From an augmented view of an image, we train the online network to predict the target network representation of the same image under a different augmented view. At the same time, we update the target network with a slow-moving average of the online network. While state-of-the art methods rely on negative pairs, BYOL achieves a new state of the art without them. BYOL reaches $74.3\%$ top-1 classification accuracy on ImageNet using a linear evaluation with a ResNet-50 architecture and $79.6\%$ with a larger ResNet. We show that BYOL performs on par or better than the current state of the art on both transfer and semi-supervised benchmarks. Our implementation and pretrained models are given on GitHub.