Data-Efficient Backdoor Attacks

Authors: Pengfei Xia, Ziqiang Li, Wei Zhang, Bin Li

IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The experimental results on CIFAR-10 and Image Net-10 indicate that the proposed method is effective: the same attack success rate can be achieved with only 47% to 75% of the poisoned sample volume compared to the random selection strategy.
Researcher Affiliation Academia University of Science and Technology of China, Hefei, China {xpengfei,iceli,zw1996}@mail.ustc.edu.cn, binli@ustc.edu.cn
Pseudocode Yes Algorithm 1: Filtering-and-Updating Strategy
Open Source Code Yes The prototype code of our method is now available at https://github. com/xpf/Data-Efficient-Backdoor-Attacks.
Open Datasets Yes We perform experiments on CIFAR-10 [Krizhevsky and Hinton, 2009] and Image Net-10 to test the effectiveness of the proposed method.
Dataset Splits No The paper refers to training and test sets but does not explicitly specify a validation dataset split (e.g., percentages or counts for a validation set).
Hardware Specification No The paper does not provide specific details about the hardware used for running the experiments (e.g., GPU models, CPU types).
Software Dependencies No The paper mentions optimizers (SGD and ADAM) but does not specify version numbers for any software dependencies.
Experiment Setup Yes The total training duration is 60, and the batch size is 512 for CIFAR-10 and 256 for Image Net-10. The initial learning rate is set to 0.01 and is dropped by 10 after 30 and 50 epochs. ... α is set to 0.5 and N is set to 10, if not otherwise specified.