Hello, you did a great job with Forgerynet.

I requested and obtained access to the Training folder, I downloaded one of the files contained, but I can't open it because of its extension. "Training.tar.4" for example, how can i open this folder as an ".tar"?

Hi, thanks for the amazing work! I checked several samples of the image subset and found that sometimes there is more than one face in an image. Will you provide the exact face bounding box coordinates for image-level detection? Or could you please provide more information about the face detector you used in this work? Thanks a lot!

发现数据集中一些数据标注与标签文件不同。 2 train_video_release/1/987080cdf7ec51e4c46c1055d6cdd0eb/47d69f53ec52643354975550e39130be/video.mp4 79 17 1 96 64 0 1 2 1 在数据集中发现在79帧以前还有一些数据是操纵过了的 但是没有标记，能解释一下数据标注吗？还是训练过程中只用标注过的数据？

Good morning. I am trying to download the training and validation set. I got access to Google Drive but I need to be able to download it via wget. I can't because authentication is required and then via wget I am redirected to the login page. I have also tried using the 105 site but I cannot register because they do not accept Italian phone numbers.

Do you have any alternatives to allow me to download via wget?

It appears that the validation (and training) set is now stored in a different location. I'm trying to download it from https://www.aliyundrive.com/s/M1FWYLZs4HJ/folder/611f0735642e1e9cf16c4e5d8d74305891be45fb, but because the site is not in English and requires an account, it was not possible to download it.

Would it be possible to move at least the validation set to somewhere else, e.g., GoogleDrive?

Hi,

I tried downloading the dataset from here https://yinanhe.github.io/projects/forgerynet.html, but I couldn't find the "Download * Set" button to click. Could you help me out please?

Many thanks!

Hello,

I'd appreciate if you could share your google drive dataset with me, my team and I are working on a Deepfake detection research project. My gmail is [email protected] . Thank you.

Hi, I think you add lots of data augmentation to the validation set. This results in a large gap between the training and validation sets. This can lead to a situation where a model performs well, possibly simply due to a closer augmentation setup used in the validation set. It is very unfair when compared methods on the validation set. So, I suggest you provide detailed data augmentation setup or codes mentioned in the paper.

Hi,I have downloaded the dataset from google drive. But when I tried to unzip the Training.tar.0 by :

tar -xvf Training.tar.0

the error comes as follow: tar:Unexpected EOF in archive tar:Unexpected EOF in archive tar: Error is not recoverable: exiting now and when I try to unzip the Training.tar.2:

Why this happened? How could I get the data? Whether the files in google drive broken or not？ Please give me some advice, thank you.

I download the model from google drive.When unzipping, a message is displayed indicating that the file is corrupted. Can you give me a new download link? For example, Baidu network disk.

I want to do some experiment to reproduce the task "Spatial Forgery Localization", but I don't know how to get the same face crop as yours while the face crop area is related to the final IoU results.

Hi,

Thanks for this excellent dataset. I would like to know how we should unzip the dataset? For the training set we get 5 files named as Training.tar.0 ... Training.tar.4.

When I unzip the files I don't get the directories containing fake data samples. I only get folders named, 18, 19 and 1.

Am I missing something here? Could you please guide me on how I should extract the data?

Thanks

Hello, thanks for sharing us such excellent work! Here i have some issues about the label list: taking a video forgery as example, "3 val_video_release/4/7f89846f0181198f2aee02c863c2aca5/718e0e03ab494625ccd35e839deafadb/video.mp4 0 10 0 10 35 1 45 30 0 1 2 4", what does these suffix numbers mean? "0 10 0 10 35 1 45 30 0 1 2 4"

Hello! You provided the weights for XceptionNet working on images. Can you please make public also the weights of SlowFast and X3D-M trained on the videos?

Thank you!

Hi,

we can open ForgeryNet in Chrome browser：https://drive.google.com/drive/folders/1gViqLc4l9sT1WTB-_hlcDUCWYskT2_LE, but I cannot download with gdrive tool, it shows ERROR as following:

Failed to get file: googleapi: Error 404: File not found: 1gViqLc4l9sT1WTB-_hlcDUCWYskT2_LE., notFound

When I try to get link of each tar file, it shows Restriced access: Only people with access can open with the link. Can you check this error and help me~ Thanks @yinanhe

Although there are videos in the training set manipulated by methods 2 and 8, these are not reported in the video_list.txt. I have added them by generating them within a new video_list.txt file, but I do not have the time segmentation information. Could you provide it?

In the meantime, if anyone needs it, here is the modified file including videos 2 and 8 with the information we have available: video_list_complete.txt