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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Although Generative Adversarial Networks (GANs) are successfully applied todiverse fields, training GANs on synthetic aperture radar (SAR) data is achallenging task mostly due to speckle noise. On the one hands, in a learningperspective of human's perception, it is natural to learn a task by usingvarious information from multiple sources. However, in the previous GAN workson SAR target image generation, the information on target classes has only beenused. Due to the backscattering characteristics of SAR image signals, theshapes and structures of SAR target images are strongly dependent on their poseangles. Nevertheless, the pose angle information has not been incorporated intosuch generative models for SAR target images. In this paper, we firstly proposea novel GAN-based multi-task learning (MTL) method for SAR target imagegeneration, called PeaceGAN that uses both pose angle and target classinformation, which makes it possible to produce SAR target images of desiredtarget classes at intended pose angles. For this, the PeaceGAN has twoadditional structures, a pose estimator and an auxiliary classifier, at theside of its discriminator to combine the pose and class information moreefficiently. In addition, the PeaceGAN is jointly learned in an end-to-endmanner as MTL with both pose angle and target class information, thus enhancingthe diversity and quality of generated SAR target images The extensiveexperiments show that taking an advantage of both pose angle and target classlearning by the proposed pose estimator and auxiliary classifier can help thePeaceGAN's generator effectively learn the distributions of SAR target imagesin the MTL framework, so that it can better generate the SAR target images moreflexibly and faithfully at intended pose angles for desired target classescompared to the recent state-of-the-art methods.

PeaceGAN: A GAN-based Multi-Task Learning Method for SAR Target Image Generation with a Pose Estimator and an Auxiliary Classifier