Backdoor

Attacks that subtly affect a model to achieve an adversarial goal, while maintaining the benign performance

---

source

BackdoorAttack

 BackdoorAttack (after_create=None, before_fit=None, before_epoch=None,
                 before_train=None, before_batch=None, after_pred=None,
                 after_loss=None, before_backward=None,
                 after_cancel_backward=None, after_backward=None,
                 before_step=None, after_cancel_step=None,
                 after_step=None, after_cancel_batch=None,
                 after_batch=None, after_cancel_train=None,
                 after_train=None, before_validate=None,
                 after_cancel_validate=None, after_validate=None,
                 after_cancel_epoch=None, after_epoch=None,
                 after_cancel_fit=None, after_fit=None)

A Callback that affects the training process to install a backdoor. Also allows the measuring of the attack’s success

---

source

BackdoorAttack._asr_dl

 BackdoorAttack._asr_dl ()

Returns a DataLoader used to measure the ASR (attack success rate)

---

source

DataPoisoningAttack

 DataPoisoningAttack (test_only=False, poison_fraction=0.1)

A BackdoorAttack that installs the backdoor by altering a small portion of the training dataset

---

source

DataPoisoningAttack._poison

 DataPoisoningAttack._poison (data_to_poison:fastai.data.core.Datasets)

	Type	Details
data_to_poison	Datasets	The prtion of the clean training data that will be replaced by poison. Could be used to construct the poison data
Returns	Datasets	A dataset of poison data