NERO-Net: A Neuroevolutionary Approach for the Design of Adversarially Robust CNNs

1. 1Utilize neuroevolution to automate neural network design
2. 2Incorporate adversarial robustness as a key objective in the design process
3. 3Evaluate the robustness of evolved models using adversarial attacks
4. 4Refine the design process based on the evaluation results

ML researchers and engineers designing neural networks for safety-critical applications can benefit from NERO-Net, as it provides a novel approach to creating intrinsically robust models

💡 Neuroevolution can be used to design CNNs with intrinsic robustness to adversarial attacks

🤖 NERO-Net: a neuroevolutionary approach for designing adversarially robust CNNs 🚀