zhen8838 / animestylized Goto Github PK

I think this line should be changed to :
'''
x = self.conv1(self.leaky_relu(self.conv(inputs)))
'''

您好，作者，我仔细比对了论文和您的源代码，我发现colorshift中，论文里的公式跟您给的计算方法不一样，而且这么看的话colorshift并没有给出一张灰度图

您好，作者，请问你的random split函数在哪里，我下了您的代码中没有看到

There are files for a different use, can you explain which of them are for model and give info about the contents of the different folder?

想请问作者数据集的来源？

首先感谢作者提供pytorch版本代码。
源码（tensorflow）里面没有直接提供训练数据，在issue里作者提供的训练数据如scenery_cartoon包含三个文件夹，shinkai，hosoda，hayao。请问训练的时候是合并在一起还是像这里面默认只选择了shinkai的数据训练？包括face_cartoon也是两个文件夹，这里也是只选择了其中一个

您好，
非常感谢您能开源代码。
在使用过程中遇到了一点小问题，想与您交流一下。

我测试的代码是《White-box Cartoon》这篇论文的复现代码。

在train的时候，会报一个错误。
这个错误的位置是whiteboxgan_pretrain.py的30-31行，看了一下dataset的构造和后序训练的输入，在30和31行之间，是否应当增加

input_photo = torch.cat(input_photo, dim=0)

否则input_photo是一个元组有两个元素，没有进行合并。

Great work in re-implementing the white-box model using pytorch. However, after testing, I found the results are not good as the official version by the authors. There is still a large gap. What do you think could be the reason? Do we need the train the model longer or something else?

!make infer CODE=scripts/whiteboxgan.py
CKPT=whitebox/version_0/checkpoints/epoch=4.ckpt
EXTRA=image_path:tests/rj.mp4,device:cuda,batch_size:4

output video has no audio channel.

File "scripts/whiteboxgan.py", line 186, in training_step
vgg_output = self.pretrained(output)
File "/usr/local/python3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/home/cyr/17_style/networks/pretrainnet.py", line 172, in forward
return self._forward_impl(x)
File "/home/cyr/17_style/networks/pretrainnet.py", line 165, in _forward_impl
x = self._process(x)
File "/home/cyr/17_style/networks/pretrainnet.py", line 156, in _process
return self.vgg_normalize(bgr) # vgg norm
File "/home/cyr/17_style/networks/pretrainnet.py", line 161, in
self.vgg_normalize = lambda x: x - mean
RuntimeError: expected device cuda:1 but got device cuda:0

配置文件中g_blur_weight和g_gray_weight是1和0.1， 代码位置

而原始whiteboxgan实现中g_blur_weight和g_gray_weight是0.1和1, 代码位置

How can I process video

这是一个超级棒的项目。感谢~
这个项目的license是什么？需要商业授权吗？