3 ms·
Adversarial attacks rely on a specific model’s gradient (since you’re essentially trying to find the most sensitive pixels). Adversarial noise that affects mod
by KMnO4 6y ago
Adversarial attacks rely on a specific model’s gradient (since you’re essentially trying to find the most sensitive pixels).
Adversarial noise that affects model A won’t necessarily work on model B. That said, most people transfer train from well trained nets (ImageNet, Inception, etc).
Finally, not all SOTA methods are susceptible to adversarial attacks, eg capsule networks.
- throw99901 6y ago>Finally, not all SOTA methods are susceptible to adversarial attacks, eg capsule networks. They appear to be susceptible: https://arxiv.org/pdf/1906.03612.pdf https://arxiv.org/pdf/1906.03612.pdf
- KMnO4 6y agoThat’s neat; hadn’t seen that paper. Thanks for sharing.