Improving Robustness of Deep Learning (Evolutionary Computation Adversarial Attack)

We propose a new Adversarial Attack combining Projected Gradient Descent and Evolutionary Computation.

avatar Samet Demir
poster
Poster

We propose a new Adversarial Attack combining Projected Gradient Descent and Evolutionary Computation. Our experiments show the method improves the baseline and it achieves near state-of-the-art.

Poster: https://drive.google.com/file/d/1c2MZ7n9q-tL3Ad3wr4i6sadp0i2V_E3l/view?usp=sharing

Video: https://drive.google.com/file/d/12TTlpAwXBm6gaaV9AoGoeqXPYyFLFxPp/view?usp=sharing

Advisors

Alper Şen