laulampaul
laulampaul
Nice work ! And could I know when will the codes be released?
I notice that in your codes, you extract the feature maps from the instance norm layer. （4,8,12,16） what happen if removing the instance norm layer ? Does patchnce loss function...
when I use the pacconvtranspose2d, there is a bug. expected object of backed CUDA but got backend CPU for argument #2 'mat2' How to deal with it? Thank you.
Excuse me, your codes don't have the 'data' folder, so where and how can I download the dataset?
in dis_update(self, x_a, x_b, hyperparameters): the content adversial codes are , out_a = self.dis_content(h_a) out_b = self.dis_content(h_b) why not use h_a.detach() and h_b.detach() to avoid the update in the parameters...
When I use the codes : ``` def calculate_cobiloss(img,gt): bb = img.shape[0] loss = 0. cobiloss = cl.ContextualBilateralLoss(use_vgg=False, loss_type='cosine').cuda() for i in range(bb): imgpatches = sample_patches(img[i],10,5) gtpatches = sample_patches(gt[i],10,5) c,...
打开该地址提示已经失效 https://bhpan.buaa.edu.cn/#/link/10A09C6A5905C5418D769693A645DAEC 其他预训练模型可以正常打开 请问能再提供一下256 bitrate这个模型的下载地址么？谢谢
Some blogs pointed out that (https://blog.csdn.net/watermelon1123/article/details/79262078) BRISQUE lower , the performance is better. Why BRISQUE is higher, the performance is better in your codes?
In code/models.py , line 11 it seems that the RIDmodel is lost. where can I get the codes? thank you!
could i ask a question about how to update the mask？cause i dont understand the code clearly. in the mask，1 means need to process and 0 means dont , in...