1 paper · 1 filter
Haoheng Lan, Jindong Gu, Philip Torr +1
When a small number of poisoned samples are injected into the training dataset of a deep neural network, the network can be induced to exhibit malicious behavior during inferences,…