gwding
commented
4 years ago Hi @MasanoriYamada Thanks for the suggestion. It would be really nice if you could contribute your implementation. As long as contribution contains a benchmark report like this https://github.com/BorealisAI/advertorch/blob/master/advertorch_examples/attack_benchmarks/benchmark_fast_adaptive_boundary.py it'll be useful for the community. and an error bar would be even better, which we can also incorporate into the common practice of advertorch.
MasanoriYamada
A simple spatial transform attack was proposed at ICML2019.
This is a simple attack that creates an adversarial example of image rotation and translation with a grid search (or random).
The advertorch already has a spatial transform attack ICLR2018 is implemented, but the simpler ICML I think adding 2019 would be useful.
If possible, I am willing to contribute to this repository and share my implementation of the method compatible with your interface later.
However, I only checked the accuracy compared with the paper using mnist. Because I only tried a random attack due to memory capacity issues, and this attack relies on random numbers
If I were to contribute, I would check for accuracy with an error bar.