DeblurGANv2

DeblurGANv2 copied to clipboard

For DVD dataset, the original authors test set contains 100 images in each folder and [here](https://github.com/shuochsu/DeepVideoDeblurring/issues/2#issuecomment-312954167) they explain that they use only 30 frames per folder for testing. However, on...

adityac8

After the model training, what files can be run to get the final test results of the author? Is PSNR in training the result?

WongJianyang

Would it be possible to provide a guide on how to deblur a video?

2

Since no instructions are provided on how to deblur a video instead of an images, would it be possible to provide guidelines for the same?.

gupta-keshav

test

6

how to get best_fpn.h5,is own train ? i just want to test the deblur result

cLEARLoVE00

RAGAN-LS loss is difference between code and paper

1

Hi~ Thanks for your work. There are some question about RAGAN-LS loss, the implementation in code is self.loss_D = (torch.mean((self.pred_real - torch.mean(self.fake_pool.query()) - 1) ** 2) + torch.mean((self.pred_fake - torch.mean(self.real_pool.query())...

xhhong

Train slowly under window

1

I ran the code under W7 GeForce RTX 2080 Ti, and met problem.I modified the code as below follow https://github.com/TAMU-VITA/DeblurGANv2/issues/29, But i find it train too slow. Does anybody know...

lisishen

x2num.py:14] NaN or Inf found in input tensor

1

use my own data, without pretrained model , x2num.py:14] NaN or Inf found in input tensor

yulinhuyang

1920*1080 image It doesn't work

1

1920*1080 image ，I use my own pictures，It doesn't work，big size image it doesn't work，I use the enlarged GoPro image，it work

lostfordream

Adding an fixed Auxiliary Classifier generate loss to the lossD or lossG part

6

hi， I am trying to do a face recognition project 。Some of my Images are motion blured baddly，my trained face recognition model mobilefacenet(this trained model is fixed ) get bad...

vgg4resnet

lenalinyajing