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

Without sufficient data, the quantity of information available for supervisedtraining is constrained, as obtaining sufficient synthetic aperture radar (SAR)training data in practice is frequently challenging. Therefore, current SARautomatic target recognition (ATR) algorithms perform poorly with limitedtraining data availability, resulting in a critical need to increase SAR ATRperformance. In this study, a new method to improve SAR ATR when training dataare limited is proposed. First, an embedded feature augmenter is designed toenhance the extracted virtual features located far away from the class center.Based on the relative distribution of the features, the algorithm pulls thecorresponding virtual features with different strengths toward thecorresponding class center. The designed augmenter increases the amount ofinformation available for supervised training and improves the separability ofthe extracted features. Second, a dynamic hierarchical-feature refiner isproposed to capture the discriminative local features of the samples. Throughdynamically generated kernels, the proposed refiner integrates thediscriminative local features of different dimensions into the global features,further enhancing the inner-class compactness and inter-class separability ofthe extracted features. The proposed method not only increases the amount ofinformation available for supervised training but also extracts thediscriminative features from the samples, resulting in superior ATR performancein problems with limited SAR training data. Experimental results on the movingand stationary target acquisition and recognition (MSTAR), OpenSARShip, andFUSAR-Ship benchmark datasets demonstrate the robustness and outstanding ATRperformance of the proposed method in response to limited SAR training data.

SAR ATR Method with Limited Training Data via an Embedded Feature Augmenter and Dynamic Hierarchical-Feature Refiner