eduzhai > Applied Sciences > Engineering >

Generative Adversarial Networks for Synthesizing InSAR Patches

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 6 pages

Abstract: Generative Adversarial Networks (GANs) have been employed with certainsuccess for image translation tasks between optical and real-valued SARintensity imagery. Applications include aiding interpretability of SAR sceneswith their optical counterparts by artificial patch generation and automaticSAR-optical scene matching. The synthesis of artificial complex-valued InSARimage stacks asks for, besides good perceptual quality, more stringent qualitymetrics like phase noise and phase coherence. This paper provides a signalprocessing model of generative CNN structures, describes effects influencingthose quality metrics and presents a mapping scheme of complex-valued data togiven CNN structures based on popular Deep Learning frameworks.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×