Real-Time Adversarial Attacks

Authors: Yuan Gong, Boyang Li, Christian Poellabauer, Yiyu Shi

IJCAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In our experiments, we test the dataset and target model mentioned in Section 3.1. The data is split as follows: we first hold out 20% of the data as the test set (test set 2) for evaluating the attack performance; so it is not seen by the target model and the attack model. We use the other 80% of the data to train the target voice recognition model; this same set is then reused to develop the attack model.
Researcher Affiliation Academia Yuan Gong , Boyang Li , Christian Poellabauer and Yiyu Shi University of Notre Dame {ygong1, bli1, cpoellab, yshi4}@nd.edu
Pseudocode Yes Algorithm 1 Real-time Adversarial Attack
Open Source Code Yes Corresponding code and demos are available at https://github. com/Yuan Gong ND/realtime-adversarial-attack
Open Datasets Yes We train the voice command recognition model exactly as in the implementation of the Tensorflow example using the voice command dataset [Warden, 2018]
Dataset Splits Yes The data is split as follows: we first hold out 20% of the data as the test set (test set 2) for evaluating the attack performance; so it is not seen by the target model and the attack model. We use the other 80% of the data to train the target voice recognition model; this same set is then reused to develop the attack model. Specifically, we use 75% of this set to train the attack model (attack training set), 6.25% for validation, and 18.75% for testing (test set 1).
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions 'TensorFlow' and 'ADAM optimizer' but does not provide specific version numbers for these or any other key software dependencies.
Experiment Setup Yes We use 1e-3 as the learning rate, mean square loss, and ADAM optimizer [Kingma and Ba, 2014] for training. [...] As shown in Table 1, we use an end-to-end neural network. [...] The frequency of the speech signal (i.e., 16 k Hz) is much higher than the possible update speed of the real-time adversarial perturbation generator. Therefore, we apply batch processing as mentioned in Section 2.3; specifically, the adversarial generator updates every 0.01 second and each update makes a decision on the actions for 0.01 seconds, so the delay is also 0.01 seconds.