deepfake_detetction_test

[njustczr]

if i use the script, i just put the test data like this?

[vishal3477]

@njustczr, yes. That is correct.

[njustczr]

@njustczr, yes. That is correct.

thanks, i also want to know if '0' is the fake image and '1' is the real image? and there are three weights in the deepfake_detection, is there any difference between them?

[vishal3477]

@njustczr , '0' is the fake image and '1' is the real image. Also, the pre-trained models are provided for the data the model is trained on. The folder name tells about the data.

[njustczr]

@njustczr , '0' is the fake image and '1' is the real image. Also, the pre-trained models are provided for the data the model is trained on. The folder name tells about the data.

i mean in the deepfake detection folder, there are three weights like the following image

[vishal3477]

@njustczr , can you tell me the link you are accessing? Because, I'm seeing this from my end.

Can you try accessing the models using https://drive.google.com/drive/folders/1bzh9Pvr7L-NyQ2Mk-TBSlSq4TkMn2anB?usp=sharing ?

[njustczr]

@njustczr , can you tell me the link you are accessing? Because, I'm seeing this from my end.

Can you try accessing the models using https://drive.google.com/drive/folders/1bzh9Pvr7L-NyQ2Mk-TBSlSq4TkMn2anB?usp=sharing ?

Yes, we are the same.But i do not know what's the difference between 'uadfv', 'dfdc_uadfv' and 'dfdc'?

[vishal3477]

So uadfv is model trained on UADFV data, dfdc is model trained on DFDC data, and dfdc_uadfv is model trained on DFDC+UADFV data combined.

[njustczr]

So uadfv is model trained on UADFV data, dfdc is model trained on DFDC data, and dfdc_uadfv is model trained on DFDC+UADFV data combined.

thanks