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

Learning representations through self-supervision on unlabeled data hasproven highly effective for understanding diverse images. However, remotesensing images often have complex and densely populated scenes with multipleland objects and no clear foreground objects. This intrinsic property generateshigh object density, resulting in false positive pairs or missing contextualinformation in self-supervised learning. To address these problems, we proposea context-enhanced masked image modeling method (CtxMIM), a simple yetefficient MIM-based self-supervised learning for remote sensing imageunderstanding. CtxMIM formulates original image patches as a reconstructivetemplate and employs a Siamese framework to operate on two sets of imagepatches. A context-enhanced generative branch is introduced to providecontextual information through context consistency constraints in thereconstruction. With the simple and elegant design, CtxMIM encourages thepre-training model to learn object-level or pixel-level features on alarge-scale dataset without specific temporal or geographical constraints.Finally, extensive experiments show that features learned by CtxMIM outperformfully supervised and state-of-the-art self-supervised learning methods onvarious downstream tasks, including land cover classification, semanticsegmentation, object detection, and instance segmentation. These resultsdemonstrate that CtxMIM learns impressive remote sensing representations withhigh generalization and transferability. Code and data will be made publicavailable.

CtxMIM: Context-Enhanced Masked Image Modeling for Remote Sensing Image Understanding