1 code implementation • 10 Aug 2020 • Gunnar Mein, Kevin Hartman, Andrew Morris
We present co-tuning with a synthetic data generator as a highly effective method to protect against 95% of pre-manufactured adversarial samples while maintaining 98% of original benchmark performance.
1 code implementation • 20 Dec 2019 • Ylenia Giarratano, Eleonora Bianchi, Calum Gray, Andrew Morris, Tom MacGillivray, Baljean Dhillon, Miguel O. Bernabeu
Our results show that, for the set of images considered, deep learning architectures (U-Net and CS-Net) achieve the best performance.