gwinhen / backdoorvault Goto Github PK

in line 205-207, I think for attack like refool. The logic of the current implementation is for front n_poison samples, randomly find a non-target sample and make it become a poisoned sample by adding triggers. Actually, when randomly selecting clean sample from other classes, till the end of the training, every clean sample can be transformed into a poisoned sample, so the actual poison rate could be 100%. By deactivate this, I find the ASR of refool_smooth and refool_ghost, the current version ASR could be 100%, while after I modify these lines, the ASR is around 90%, I think it is reasonable. If you have question, we can discuss.

if args.attack == 'dfst_detox'
i can't find where you use the unet you save in detox when you train the model
can you explain it ?