… contrastive learning of visual representations. We simplify recently proposed contrastive
selfsupervised learning … enables the contrastive prediction tasks to learn useful representations, …

… Our objective is to learn visual representation from multimodal data in a unified training
framework. … We aim to learn high-quality visual features, which benefit from not only the additional …

… representation learning techniques have largely closed the gap between supervised and
unsupervised learning … direction, we study contrastive selfsupervised learning on four diverse …

… that make representations from contrastive learning more transferable … representations of
contrastive learning are much different than the cross-entropy model; in particular, contrastive …

… contrastive learning method SimCLR [7] and in section 4.2 we introduce attempts to improve
contrastive … Section 5.1 presents the model bias [44] in contrastive learning, and is followed …

… learning method called Mutual Contrastive Learning (MCL) for general visual representation
learning… The core idea of MCL is to perform mutual interaction and transfer of contrastive …

… -supervised visual representation learning … contrastive visual representation learning for
videos and images. The method, Video/Image for Visual Contrastive Learning of Representation …

… a viable alternative to learn dense representations without relying on video data [24… is to
learn more effective representations through contrastive self-supervised learning without relying …

… To validate the efficacy of MixCo, it is applied to recent popular representation learning …
extended the concept of contrastive loss for visual representation learning to semi-positive …

… In this work, we modify SimCLR [12], a state-of-the-art contrastive representation method,
by adding information compression using the Conditional Entropy Bottleneck (CEB) [27]. …