Gated Self-supervised Learning For Improving Supervised Learning
This work addresses a bottleneck in self-supervised learning for image classification researchers, offering an incremental improvement over existing methods.
The paper tackles the problem of limited feature learning in self-supervised image classification by proposing a novel approach that combines multiple augmentations with a gating method, resulting in improved performance on classification tasks.
In past research on self-supervised learning for image classification, the use of rotation as an augmentation has been common. However, relying solely on rotation as a self-supervised transformation can limit the ability of the model to learn rich features from the data. In this paper, we propose a novel approach to self-supervised learning for image classification using several localizable augmentations with the combination of the gating method. Our approach uses flip and shuffle channel augmentations in addition to the rotation, allowing the model to learn rich features from the data. Furthermore, the gated mixture network is used to weigh the effects of each self-supervised learning on the loss function, allowing the model to focus on the most relevant transformations for classification.