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

Labeling semantic segmentation datasets is a costly and laborious process ifcompared with tasks like image classification and object detection. This isespecially true for remote sensing applications that not only work withextremely high spatial resolution data but also commonly require the knowledgeof experts of the area to perform the manual labeling. Data augmentationtechniques help to improve deep learning models under the circumstance of fewand imbalanced labeled samples. In this work, we propose a novel dataaugmentation method focused on exploring the spatial context of remote sensingsemantic segmentation. This method, ChessMix, creates new synthetic images fromthe existing training set by mixing transformed mini-patches across the datasetin a chessboard-like grid. ChessMix prioritizes patches with more examples ofthe rarest classes to alleviate the imbalance problems. The results in threediverse well-known remote sensing datasets show that this is a promisingapproach that helps to improve the networks' performance, working especiallywell in datasets with few available data. The results also show that ChessMixis capable of improving the segmentation of objects with few labeled pixelswhen compared to the most common data augmentation methods widely used.

ChessMix: Spatial Context Data Augmentation for Remote Sensing Semantic Segmentation