search

mapooon / SelfBlendedImages

[CVPR 2022 Oral] Detecting Deepfakes with Self-Blended Images https://arxiv.org/abs/2204.08376
Other
185 stars 32 forks source link

The AUC of CelebDF #7

Closed JieLiu95 closed 2 years ago

JieLiu95 commented 2 years ago

I run the inference on Celeb-DF-v2 with your pre trained model, but the AUC is 0.9235, which is lower than yours. And the versions of torch and torchvision are the same with you. Do you know the reason? Can you provide the version of opencv?

mapooon commented 2 years ago

The version of opencv (opencv-python) is 4.5.4.60. It is strongly recommended to use the provided docker file to reproduce the experimental results.

JieLiu95 commented 2 years ago

Thanks a lot. I have solved it when I change the version of efficientnet-pytorch.

YU-SHAO-XU commented 1 year ago

@JieLiu95 Can you provide the version of efficientnet-pytorch!?

JieLiu95 commented 1 year ago

0.7.1 @YU-SHAO-XU

YU-SHAO-XU commented 1 year ago

@JieLiu95 and your SBI (train) is 1421 !? i am sure that landmark augmentation is used because show 'exist library' . I train from scratch and test on CDF AUC only 90.24, so i wanna check if development environment or datst have problem thank you!

YU-SHAO-XU commented 1 year ago

@JieLiu95 did u train from scratch ? or only use weight that author provide?!

JieLiu95 commented 1 year ago

@JieLiu95 and your SBI (train) is 1421 !? i am sure that landmark augmentation is used because show 'exist library' . I train from scratch and test on CDF AUC only 90.24, so i wanna check if development environment or datst have problem thank you!

I train it from scratch. Did you utilize the mask of X-ray as illustrated in Step 4. mkdir src/utils/library git clone https://github.com/AlgoHunt/Face-Xray.git src/utils/library @YU-SHAO-XU

YU-SHAO-XU commented 1 year ago

@JieLiu95
is that bi_online_generation.py file ? if use that file AUC will 77.xx increase to 90.xx, or you mean DeepFakeMask.py !? and how to use DeepFakeMask !?

JieLiu95 commented 1 year ago

If you use bi_online_generation.py, I guess you have used DeepFakeMask.py by the function random_get_hull.

YU-SHAO-XU commented 1 year ago

@JieLiu95 so i think it seems to dataset or development environment ~ thank you !

YU-SHAO-XU commented 1 year ago

@JieLiu95 did u hvae error like :

  1. (-215:Assertion failed) !ssize.empty() in function 'resize'
  2. [W pthreadpool-cpp.cc:90] Warning: Leaking Caffe2 thread-pool after fork. (function pthreadpool) is it normal !?