Hello, thank you for releasing code and pretrained model. I tried to apply the model on standard face swap dataset FF++. However, although I precisely followed the procedure written on readme, results contain much more visible artifacts in substantial probability. I attached three images, but much more occurred. Do you have any idea what makes it such difficult to apply the model on FF++?

您好！
感谢这篇高质量的工作以及靠谱的开源！

我在提供的CelebA-HQ上获取了很好的复现结果。但是，在其他图片上进行测试时，lmk和segment map我可以正常提取并可视化得到不错的结果，对于code（也就是W+ latent space embedding），因为pSp提供了多个checkpoint，似乎没有指明该使用哪个？请问能否提供相关信息。我目前获取encoder的输出，即code，测试了 with&without average latent code的结果，看来都是错误的。

提供的数据集的结果：

我自己提取code/landmark/segment map的结果：

感谢！

from criteria.lpips.lpips import LPIPS
from criteria import w_norm
The error reported directly is, ModuleNotFoundError: No module named 'criteria',
Directly! pip install criteria, but found that the root quilt does not have this dependency, and the corresponding criteria.py file was not found in the source file. Where should I download and import this dependency?
Sorry to trouble you, if you see it, thank you for your reply!

is it correct to get video face swapping by inferring video frame by frame?

Hi,
have you tested your released model on FF++? I follewed all your preprocessing steps, but the synthetic inner face is about a mess.
However, following the same steps, I got proper synthetic faces on CelebAHQ dataset.
I really wonder what's wrong with FF++, and I'm looking forward to any reply with many thanks!

Hello, when I was downloading the CelebA-HQ dataset and checkpoint from the project homepage, the link became invalid. Could you please update the download link again.

Your papar mentions that the model can be extended to video face swapping with two extra losses. But it seems that the losses require video training datasets.

Thanks for your excellent work!

I recently saw this article on arxiv and would like to ask you for some details.

1. At the "Background Transfer" part. there has a sentence like:

We then aggregate each pair of corresponding features (fis, fti) by replacing the components of fti for the inner-face region with their counterparts in fis.
I want to know how this region-wise replacement do?

2. Landmark encoder use the "Encoding in Style" methods, and your input to this network is landmak points (eg. 68*2 ) or landmark picture(eg. picture plot with landmark)? how two inputs can generate face shap lantent direction $\overrightarrow{n}$.

Looking for your reply, many thx!

I tried to calculate the ID retrieval for the code using the pretrained arcface model but got the values very low. I believe my approach to calculate ID retrieval is wrong. I tried with 500 images.

during setting up the virtual env I get multiple error like: ERROR: Could not find a version that satisfies the requirement XY with the following dependencies:

• channelnorm-cuda==0.0.0
• clip==1.0
• correlation-cuda==0.0.0
• insightface==0.6.2
• resample2d-cuda==0.0.0
• torch==1.9.0+cu111

I just commented all of those out but then when I try to run bash run.sh
I get ModuleNotFoundError: No module named 'criteria'

here is the whole log:
home/ubuntu/miniconda3/envs/stylegan/lib/python3.7/site-packages/torch/distributed/launch.py:186: FutureWarning: The module torch.distributed.launch is deprecated and will be removed in future. Use torchrun. Note that --use_env is set by default in torchrun. If your script expects --local_rankargument to be set, please change it to read fromos.environ['LOCAL_RANK']` instead. See
https://pytorch.org/docs/stable/distributed.html#launch-utility for
further instructions

FutureWarning,
Traceback (most recent call last):
File "test.py", line 22, in
from utils import *
File "/home/ubuntu/FSLSD_HiRes/utils.py", line 19, in
from criteria.lpips.lpips import LPIPS
ModuleNotFoundError: No module named 'criteria'
ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: 1) local_rank: 0 (pid: 884771) of binary: /home/ubuntu/miniconda3/envs/stylegan/bin/python
Traceback (most recent call last):
File "/home/ubuntu/miniconda3/envs/stylegan/lib/python3.7/runpy.py", line 193, in _run_module_as_main
"main", mod_spec)
File "/home/ubuntu/miniconda3/envs/stylegan/lib/python3.7/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/home/ubuntu/miniconda3/envs/stylegan/lib/python3.7/site-packages/torch/distributed/launch.py", line 193, in
main()
File "/home/ubuntu/miniconda3/envs/stylegan/lib/python3.7/site-packages/torch/distributed/launch.py", line 189, in main
launch(args)
File "/home/ubuntu/miniconda3/envs/stylegan/lib/python3.7/site-packages/torch/distributed/launch.py", line 174, in launch
run(args)
File "/home/ubuntu/miniconda3/envs/stylegan/lib/python3.7/site-packages/torch/distributed/run.py", line 713, in run
)(*cmd_args)
File "/home/ubuntu/miniconda3/envs/stylegan/lib/python3.7/site-packages/torch/distributed/launcher/api.py", line 131, in call
return launch_agent(self._config, self._entrypoint, list(args))
File "/home/ubuntu/miniconda3/envs/stylegan/lib/python3.7/site-packages/torch/distributed/launcher/api.py", line 261, in launch_agent
failures=result.failures,
torch.distributed.elastic.multiprocessing.errors.ChildFailedError:

test.py FAILED

Failures:
<NO_OTHER_FAILURES>

Root Cause (first observed failure):
[0]:
time : 2023-11-02_19:12:54
host : 132-145-210-225
rank : 0 (local_rank: 0)
exitcode : 1 (pid: 884771)
error_file: <N/A>
traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html`