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

Few-shot segmentation (FSS) is proposed to segment unknown class targets withjust a few annotated samples. Most current FSS methods follow the paradigm ofmining the semantics from the support images to guide the query imagesegmentation. However, such a pattern of `learning from others' struggles tohandle the extreme intra-class variation, preventing FSS from being directlygeneralized to remote sensing scenes. To bridge the gap of intra-classvariance, we develop a Dual-Mining network named DMNet for cross-image miningand self-mining, meaning that it no longer focuses solely on support images butpays more attention to the query image itself. Specifically, we propose aClass-public Region Mining (CPRM) module to effectively suppress irrelevantfeature pollution by capturing the common semantics between the support-queryimage pair. The Class-specific Region Mining (CSRM) module is then proposed tocontinuously mine the class-specific semantics of the query image itself in a`filtering' and `purifying' manner. In addition, to prevent the co-existence ofmultiple classes in remote sensing scenes from exacerbating the collapse of FSSgeneralization, we also propose a new Known-class Meta Suppressor (KMS) moduleto suppress the activation of known-class objects in the sample. Extensiveexperiments on the iSAID and LoveDA remote sensing datasets have demonstratedthat our method sets the state-of-the-art with a minimum number of modelparameters. Significantly, our model with the backbone of Resnet-50 achievesthe mIoU of 49.58% and 51.34% on iSAID under 1-shot and 5-shot settings,outperforming the state-of-the-art method by 1.8% and 1.12%, respectively. Thecode is publicly available at https://github.com/HanboBizl/DMNet.

Not Just Learning from Others but Relying on Yourself: A New Perspective on Few-Shot Segmentation in Remote Sensing