search

Vchitect / Latte

Latte: Latent Diffusion Transformer for Video Generation.
Apache License 2.0
1.44k stars 147 forks source link

FaceForensics数据集 #67

Open likeatingcake opened 3 months ago

likeatingcake commented 3 months ago

现在好像无法下载FaceForensics数据集，只可以下载FaceForensics++，请问使用FaceForensics++训练模型可以吗

maxin-cn commented 3 months ago

现在好像无法下载FaceForensics数据集，只可以下载FaceForensics++，请问使用FaceForensics++训练模型可以吗

You can use any dataset you want. However, if you want to replace FaceForensics with Faceforensics ++, I'm not so sure the two datasets are the same.

ShunLu91 commented 1 month ago

现在好像无法下载FaceForensics数据集，只可以下载FaceForensics++，请问使用FaceForensics++训练模型可以吗

I have consulted the organization of FaceForensics. After application, you can download the FaceForensics using '—server EU2 flag' from the script of FaceForensics++, i.e. using the 'SERVER_URL' as below:

SERVER_URL = 'http://kaldir.vc.in.tum.de/faceforensics/'