English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Contrastive learning predicts whether two images belong to the same categoryby training a model to make their feature representations as close or as faraway as possible. In this paper, we rethink how to mine samples in contrastivelearning, unlike other methods, our approach is more comprehensive, taking intoaccount both positive and negative samples, and mining potential samples fromtwo aspects: First, for positive samples, we consider both the augmented sampleviews obtained by data augmentation and the mined sample views through datamining. Then, we weight and combine them using both soft and hard weightingstrategies. Second, considering the existence of uninformative negative samplesand false negative samples in the negative samples, we analyze the negativesamples from the gradient perspective and finally mine negative samples thatare neither too hard nor too easy as potential negative samples, i.e., thosenegative samples that are close to positive samples. The experiments show theobvious advantages of our method compared with some traditional self-supervisedmethods. Our method achieves 88.57%, 61.10%, and 36.69% top-1 accuracy onCIFAR10, CIFAR100, and TinyImagenet, respectively.

Rethinking Samples Selection for Contrastive Learning: Mining of Potential Samples