search

dongheehand / DeblurGAN-tf

Tensorflow implementation of DeblurGAN(Blind Motion Deblurring Using Conditional Adversarial Networks)
80 stars 24 forks source link

Implemented resblock is different from the original paper #8

Closed hanzotruongdev closed 4 years ago

hanzotruongdev commented 5 years ago

The Res Block has only 1 convolution as mentioned in the original paper but your implementation, your Res Block has 2 convolutional layers. Why there is this difference?

deblurgan-arch

If there is any difference, could you please include it in your document?

dongheehand commented 5 years ago

My implementation is based on author's official code and fig3 in the original paper. This is the official reference code when i implement my DeblurGAN code. ( https://github.com/KupynOrest/DeblurGAN/blob/dd9e940f92db10f0e5ce073344159b99b22dbfa0/models/networks.py )

But the official code has been modified, I am going to modify my code.

Thank you for letting me know!

hanzotruongdev commented 5 years ago

The author's office code used 2 convolutions and there is a dropout layer between them. Anyway, your code works perfectly Thank