search

HongguLiu / Deepfake-Detection

The Pytorch implemention of Deepfake Detection based on Faceforensics++
https://github.com/ondyari/FaceForensics
Apache License 2.0
311 stars 61 forks source link

Test the video with your own training model #20

Open Nemesis0416 opened 3 years ago

Nemesis0416 commented 3 years ago

Hello, I trained your pre-training model with the pictures extracted from my 6000+6000 videos, and got a high accuracy rate, but I used the trained model to test the real video, almost all the pictures in the frame The display is a fake picture. Which step has the problem occurred?

Nemesis0416 commented 3 years ago

Could you provide the literature of this detection method?

HongguLiu commented 3 years ago

How about the details of 6000 + 6000 videos? Are there 6000 fake videos and 6000 real videos?

Nemesis0416 commented 3 years ago

After the video is divided into frames, the face and surrounding parts are cropped out, and 6000 authentic and fake pictures are selected, and the number is strictly in accordance with 6000+6000.

Nemesis0416 commented 3 years ago

There are 12,000 pictures and about 300 videos.