1. Generalisation in humans and deep neural networks;geirhos;Advances in neural information processing systems,2018
2. mixup: Beyond empirical risk minimization;zhang;ICLRE,2018
3. ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness;geirhos;International Conference on Learning Representations,0
4. Wide residual networks;zagoruyko;ArXiv Preprint,2016
5. Improving Robustness using Generated Data;gowal;ArXiv Preprint,2021