search

nii-yamagishilab / ClassNSeg

Implementation and demonstration of the paper: Multi-task Learning for Detecting and Segmenting Manipulated Facial Images and Videos
BSD 3-Clause "New" or "Revised" License
78 stars 12 forks source link

class folder #6

Open Venu-ka opened 2 years ago

Venu-ka commented 2 years ago

how to solve about FileNotFoundError: Couldn't find any class folder in ./ps_data/data.

honghuy127 commented 2 years ago

The scripts use torchvision.datasets.ImageFolder dataset, therefore you need to follow its structure. Please check my instruction here and find our more about the ImageFolder here.

Venu-ka commented 2 years ago

thank you for reply！ i have another question，how to use the pre-trained models to detect a fake picture and visualize them. thank for you very much!!!

honghuy127 commented 2 years ago

Hi, there are some ways to visualize the network. You can have a look at here. For the Capsule-Forensics network, because its architecture is different from the standard CNNs, you need to modify the example's code a little bit.