Coldog2333/pytoflow

Question About Super Resolution

delldu opened this issue · 1 comments

Dear friend,
First all, we hope express our thanks for your hard work.
Second, we have an question: Why super resolution output size is same as input images ?
Best Regards,

Hello, thank you for your interest.

Actually, I upsample the original input images first. Then input them into our network. Finally, it generates the output image of super-resolution.

For example, the original input images have shapes of (batch_size, 256, 256, 3). We will firstly upsample them to (bs, 512, 512, 3), then generate (bs, 512, 512, 3) images.

Hope this helps.