jarrentwu1031 / ccpl Goto Github PK

very good work.
there's a minor error near line 41 to 47, _calc_feat_flatten_mean_std is missed.
source_f, source_f_mean, source_f_std = _calc_feat_flatten_mean_std(source)
source_f_norm = (source_f - source_f_mean.expand_as(
source_f)) / source_f_std.expand_as(source_f)
source_f_cov_eye =
torch.mm(source_f_norm, source_f_norm.t()) + torch.eye(3)

target_f, target_f_mean, target_f_std = _calc_feat_flatten_mean_std(target)

Hi, it seems that alpha is not used in the test_video_frame.py script

Hi,

I trying to execute train.py, and scripts is asking for GPU parameter,
as I am using Nvidia, so which GPU i should mention.

It is normal to load VGG and Decoder, but this error is reported when loading SCT. Please, tell me why?
RuntimeError: Error(s) in loading state_dict for SCT:
Unexpected key(s) in state_dict: "cnet.4.weight", "cnet.4.bias", "snet.4.weight", "snet.4.bias".
size mismatch for cnet.0.weight: copying a param with shape torch.Size([256, 512, 1, 1]) from checkpoint, the shape in current model is torch.Size([128, 256, 1, 1]).
size mismatch for cnet.0.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for cnet.2.weight: copying a param with shape torch.Size([128, 256, 1, 1]) from checkpoint, the shape in current model is torch.Size([32, 128, 1, 1]).
size mismatch for cnet.2.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for snet.0.weight: copying a param with shape torch.Size([256, 512, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 256, 3, 3]).
size mismatch for snet.0.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for snet.2.weight: copying a param with shape torch.Size([128, 256, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 128, 1, 1]).
size mismatch for snet.2.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for uncompress.weight: copying a param with shape torch.Size([512, 32, 1, 1]) from checkpoint, the shape in current model is torch.Size([256, 32, 1, 1]).
size mismatch for uncompress.bias: copying a param with shape torch.Size([512]) from checkpoint, the shape in current model is torch.Size([256]).

Hi, are the pre-trained models released here ready to be used for any customised content/style image pairs? Thanks!

你好，国内无法翻墙，怎么获取vgg_normalised.pth？可以给个网盘连接吗？

Hi, thank you for your greate work! I find that your code was not normalized with the pre trained ImageNet value, as shown in the title.
So I add this to the code, but when I use the trained weight to test, the performance is not good.

In contrast, I remove regularization and get the same result as yours.

So why this difference？ Can you tell me the reason? thank you!

style = coral(style, content)

Hi, I met this error when running test_video_frame.py:

Traceback (most recent call last):
  File "test_video_frame.py", line 120, in <module>
    SCT.load_state_dict(torch.load(args.SCT))
  File "D:\Anaconda\envs\CCPL\lib\site-packages\torch\nn\modules\module.py", line 1407, in load_state_dict
    self.__class__.__name__, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for SCT:
        Unexpected key(s) in state_dict: "cnet.4.weight", "cnet.4.bias", "snet.4.weight", "snet.4.bias".
        size mismatch for cnet.0.weight: copying a param with shape torch.Size([256, 512, 1, 1]) from checkpoint, the shape in current model is torch.Size([128, 256, 1, 1]).
        size mismatch for cnet.0.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([128]).    
        size mismatch for cnet.2.weight: copying a param with shape torch.Size([128, 256, 1, 1]) from checkpoint, the shape in current model is torch.Size([32, 128, 1, 1]).
        size mismatch for cnet.2.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([32]).     
        size mismatch for snet.0.weight: copying a param with shape torch.Size([256, 512, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 256, 3, 3]).
        size mismatch for snet.0.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([128]).    
        size mismatch for snet.2.weight: copying a param with shape torch.Size([128, 256, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 128, 1, 1]).
        size mismatch for snet.2.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([32]).     
        size mismatch for uncompress.weight: copying a param with shape torch.Size([512, 32, 1, 1]) from checkpoint, the shape in current model is torch.Size([256, 32, 1, 1]).
        size mismatch for uncompress.bias: copying a param with shape torch.Size([512]) from checkpoint, the shape in current model is torch.Size([256]).

It seems like something went wrong when loading the pretrained SCT, so how can I fix it?

Hi! Thanks for sharing the great work!
Could you please share more on the test video for evaluation or share the public URL link of test video for downloading?
Thanks!

I tried to untar your pre-trained model, but some errors occurred as follows

tar: This does not look like a tar archive
tar: Skipping to the next header
tar: Exiting with failure status due to previous errors

model file: sct_iter_160000.pth.tar
So how to untar this file ?

Hello, I would like to read the supplementary file of the paper but I can't find it anywhere, could you please help me?

Hi,

When I try to run the following script:
python test.py --content input/content/lenna.jpg --style input/style/in2.jpg --decoder artistic/decoder_iter_160000.pth.tar --SCT artistic/sct_iter_160000.pth.tar --testing_mode artistic

I got some errors:
File "test.py", line 128, in
SCT.load_state_dict(torch.load(args.SCT))
File "/home/huiqin/anaconda3/envs/torch18/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1604, in load_state_dict
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for SCT:
Unexpected key(s) in state_dict: "cnet.4.weight", "cnet.4.bias", "snet.4.weight", "snet.4.bias".
size mismatch for cnet.0.weight: copying a param with shape torch.Size([256, 512, 1, 1]) from checkpoint, the shape in current model is torch.Size([128, 256, 1, 1]).
size mismatch for cnet.0.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for cnet.2.weight: copying a param with shape torch.Size([128, 256, 1, 1]) from checkpoint, the shape in current model is torch.Size([32, 128, 1, 1]).
size mismatch for cnet.2.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for snet.0.weight: copying a param with shape torch.Size([256, 512, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 256, 3, 3]).
size mismatch for snet.0.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for snet.2.weight: copying a param with shape torch.Size([128, 256, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 128, 1, 1]).
size mismatch for snet.2.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for uncompress.weight: copying a param with shape torch.Size([512, 32, 1, 1]) from checkpoint, the shape in current model is torch.Size([256, 32, 1, 1]).
size mismatch for uncompress.bias: copying a param with shape torch.Size([512]) from checkpoint, the shape in current model is torch.Size([256]).

First of all, I'm glad to see your amazing works,，Can your model apply to the conversion of the style of underwater images to that of land? If I want to do so, what do I need to prepare? For example, I now have a ready-made underwater fish data set

I get message "File is in owner's thrash" from Google drive.

in pypi or conda, there are none of channelnorm-cuda. Could you tell me how to install it? thx!

Hi, congrats for the acceptance at ECCV 2022. We are having an event on Hugging Face for ECCV 2022, where you can submit spaces(web demos), models, and datasets for papers for a chance to win prizes. The hub offers free hosting and would make your work more accessible to the rest of the community. Hugging Hub works similar to github where you can push to user profiles or organization accounts, you can add the models/datasets and spaces to this organization:

https://huggingface.co/ECCV2022

after joining the organization using this link: https://huggingface.co/organizations/ECCV2022/share/kZuMIwRJKOTteDgoueNuPAMUGSnfDjWAGq

let me know if you need any help with the above steps, thanks

Hi, Thanks for your great work! Would you please provide the script about the SIFID and LPIPS that mentioned in paper?

Can you give me *.py file for extracting frames from video

Hi ,great work! And do you have plan to upload the 1024res pretrained model?