CoViews: Adaptive Augmentation Using Cooperative Views for Enhanced Contrastive Learning

TL;DR

This paper introduces CoViews, an adaptive augmentation framework that learns view-dependent policies during contrastive learning, improving representation quality without extensive supervision or differentiable augmentation generators.

Contribution

It proposes a novel adaptive augmentation method with view-dependent policies, enhancing contrastive learning by learning dependencies between transformations for each view.

Findings

CoViews outperforms baseline augmentation methods across multiple datasets.

View-dependent augmentation policies yield better representations than shared policies.

The approach requires minimal computational overhead during training.

Abstract

Data augmentation plays a critical role in generating high-quality positive and negative pairs necessary for effective contrastive learning. However, common practices involve using a single augmentation policy repeatedly to generate multiple views, potentially leading to inefficient training pairs due to a lack of cooperation between views. Furthermore, to find the optimal set of augmentations, many existing methods require extensive supervised evaluation, overlooking the evolving nature of the model that may require different augmentations throughout the training. Other approaches train differentiable augmentation generators, thus limiting the use of non-differentiable transformation functions from the literature. In this paper, we address these challenges by proposing a framework for learning efficient adaptive data augmentation policies for contrastive learning with minimal computational overhead. Our approach continuously generates new data augmentation policies during training and produces effective positives/negatives without any supervision. Within this framework, we present two methods: \ac{IndepViews}, which generates augmentation policies used across all views, and \ac{CoViews}, which generates dependent augmentation policies for each view. This enables us to learn dependencies between the transformations applied to each view and ensures that the augmentation strategies applied to different views complement each other, leading to more meaningful and discriminative representations. Through extensive experimentation on multiple datasets and contrastive learning frameworks, we demonstrate that our method consistently outperforms baseline solutions and that training with a view-dependent augmentation policy outperforms training with an independent policy shared across views, showcasing its effectiveness in enhancing contrastive learning performance.