Pinned Loading

1. Input-Significance-Indicator-based-Attack Input-Significance-Indicator-based-Attack Public

   A fast and efficient white-box iterative adversarial attack against deep learning models.

   Jupyter Notebook 1 1

2. Gradient-Projection-Regularization Gradient-Projection-Regularization Public

   Improving model's robustness to transfer attacks by regularizing projection of input gradients.

   Python 1