suke27

For SR case, it seems need calculate loss for backward, it mean it must need GT, how can i run SR in case of only LR images? is it possible?

it seems super-resolution only include x4 and x8 case, how to run less than x4 SR case

Traing dataset

1

Hi, I am very interesting in image de noise, could you share the training dataset for jpeg artifact removal?

cyclegan training

3

Hi, I have a problem for cycle gan training, currently our training dataset need many image(both A and B), can I do art style transfer based just one image. like...

we want to check whether it can reach to pytorch version. could you provide your pretrained models

problem

1

for Loss Functions for Neural Networks for Image Processing it include a*ms_ssim+(1-a)*L1*G do you realize ms_ssim only?

Queations about Mono.Data.Sqlite encryption

1

Hi, we are now use Mono.Data.Sqlite to manage DB, i have a question that, does Mono.Data.Sqlite support database encryption. is it free? i heard that it is charged. hope to...

Hi, I try to introduce scale parameter in resblock. for example ResBlock(x) { tmp = x x = conv1(x) x = conv2(x) x = x*0.1 # if i add this...