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

This paper presents Contrastive Transformer, a contrastive learning schemeusing the Transformer innate patches. Contrastive Transformer enables existingcontrastive learning techniques, often used for image classification, tobenefit dense downstream prediction tasks such as semantic segmentation. Thescheme performs supervised patch-level contrastive learning, selecting thepatches based on the ground truth mask, subsequently used for hard-negative andhard-positive sampling. The scheme applies to all vision-transformerarchitectures, is easy to implement, and introduces minimal additional memoryfootprint. Additionally, the scheme removes the need for huge batch sizes, aseach patch is treated as an image.  We apply and test Contrastive Transformer for the case of aerial imagesegmentation, known for low-resolution data, large class imbalance, and similarsemantic classes. We perform extensive experiments to show the efficacy of theContrastive Transformer scheme on the ISPRS Potsdam aerial image segmentationdataset. Additionally, we show the generalizability of our scheme by applyingit to multiple inherently different Transformer architectures. Ultimately, theresults show a consistent increase in mean IoU across all classes.

A Contrastive Learning Scheme with Transformer Innate Patches