FAPA: Transferable Adversarial Attacks Based on Foreground Attention

<div>Visualization of randomly selected clean images and corresponding adversarial images produced by our FAPA on Visformer-S model. (a) Clean. (b) Adversarial.</div>

Security and Communication Networks

fig2

Figure 2

Figure 2: FAPA: Transferable Adversarial Attacks Based on Foreground Attention 