Can this network take intensity images instead of complex images as input?

diegovalsesia / speckle2void

Speckle2Void: Deep Self-Supervised SAR Despeckling with Blind-Spot Convolutional Neural Networks

55 stars 15 forks source link

Open xiaozhuqian opened 4 months ago

xiaozhuqian commented 4 months ago

Thank you for providing the source code.

I'm wondering if I can train the network on intensity images instead of complex images, because I only have the noisy intensity images.

Thank you.

KiranRaj-B commented 4 months ago

Do you able to solve dependencies errors

xiaozhuqian commented 4 months ago

Do you able to solve dependencies errors

Yes, sort of. There were still some warnings, but they did not impact the running.