How do I give a piared-input to train a GAN for Image-to-Image Translation application?

I've been trying to change create a conditional GAN for image-to-image translation applications and I've been following several examples in the MATLAB depository. I could not find any examples related to paired-inputs/outputs or I may have missed them. Could someone direct me to an example where the GAN uses paired input/output for training?

Answer 1

[Avadhoot]

Hi @Vimalathithan PK,

You can find the documentation on using GANs for image-to-image translation on the following link:

https://www.mathworks.com/help/images/get-started-with-gans-for-image-to-image-translation.html

Additionally you can find two examples for the same in the below documentaiton links:

1. Image generation from segmentation map : https://www.mathworks.com/help/vision/ug/generate-image-from-segmentation-map-using-deep-learning.html
2. Day to dusk image translation : https://www.mathworks.com/help/images/unsupervised-day-to-dusk-image-translation-using-unit.html

I hope it helps.