Understand and Improve Contrastive Learning Methods for Visual Representation: A Review

Multilevel contrast strategy for unpaired image-to-image translation

“…In addition, some works 30,[49][50][51][52][53][54][55] have mathematically derived and experimentally demonstrated that incorporating appropriately hard negative keys can improve performance. These studies have proposed their respective methods to utilize hard negative keys effectively.…”

Hyperspectral Image Classification With Contrastive Graph Convolutional Network

“…Contrastive learning is one of self-supervised representation learning methods, which aims to obtain discriminative feature representation by leveraging the similarities and dissimilarities between examples [26]. It has gained increasing attention due to the promising results in various fields such as computer vision [7], [8] and natural language processing [27].…”