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

As a specific semantic segmentation task, aerial imagery segmentation hasbeen widely employed in high spatial resolution (HSR) remote sensing imagesunderstanding. Besides common issues (e.g. large scale variation) faced bygeneral semantic segmentation tasks, aerial imagery segmentation has someunique challenges, the most critical one among which lies inforeground-background imbalance. There have been some recent efforts thatattempt to address this issue by proposing sophisticated neural networkarchitectures, since they can be used to extract informative multi-scalefeature representations and increase the discrimination of object boundaries.Nevertheless, many of them merely utilize those multi-scale representations inad-hoc measures but disregard the fact that the semantic meaning of objectswith various sizes could be better identified via receptive fields of diverseranges. In this paper, we propose Adaptive Focus Framework (AF$_2$), whichadopts a hierarchical segmentation procedure and focuses on adaptivelyutilizing multi-scale representations generated by widely adopted neuralnetwork architectures. Particularly, a learnable module, called AdaptiveConfidence Mechanism (ACM), is proposed to determine which scale ofrepresentation should be used for the segmentation of different objects.Comprehensive experiments show that AF$_2$ has significantly improved theaccuracy on three widely used aerial benchmarks, as fast as the mainstreammethod.

AF$_2$: Adaptive Focus Framework for Aerial Imagery Segmentation