Deep neural networks (DNNs) are vulnerable to adversarial examples where inputs with imperceptible perturbations mislead DNNs to incorrect results. Recently, adversarial patch, with noise confined to a small and localized patch, emerged for its easy accessibility in real-world. However, existing attack strategies are still far from generating visually natural patches with strong attacking ability, since they often ignore the perceptual sensitivity of the attacked network to the adversarial patch, including both the correlations with the image context and the visual attention. To address this problem, this paper proposes a perceptual-sensitive generative adversarial network (PS-GAN) that can simultaneously enhance the visual fidelity and the attacking ability for the adversarial patch. To improve the visual fidelity, we treat the patch generation as a patch-to-patch translation via an adversarial process, feeding any types of seed patch and outputting the similar adversarial patch with high perceptual correlation with the attacked image. To further enhance the attacking ability, an attention mechanism coupled with adversarial generation is introduced to predict the critical attacking areas for placing the patches, which can help producing more realistic and aggressive patches. Extensive experiments under semi-whitebox and black-box settings on two large-scale datasets GTSRB and ImageNet demonstrate that the proposed PS-GAN outperforms state-of-the-art adversarial patch attack methods.

The recently described role of RNA methylation in regulating immune cell infiltration into tumors has attracted interest, given its potential impact on immunotherapy response. YTHDF1 is a versatile and powerful m6A reader, but the understanding of its impact on immune evasion is limited. Here, we reveal that tumor-intrinsic YTHDF1 drives immune evasion and immune checkpoint inhibitor (ICI) resistance. Additionally, YTHDF1 deficiency converts cold tumors into responsive hot tumors, which improves ICI efficacy. Mechanistically, YTHDF1 deficiency inhibits the translation of lysosomal genes and limits lysosomal proteolysis of the major histocompatibility complex class I (MHC-I) and antigens, ultimately restoring tumor immune surveillance. In addition, we design a system for exosome-mediated CRISPR/Cas9 delivery to target YTHDF1 in vivo, resulting in YTHDF1 depletion and antitumor activity. Our findings elucidate the role of tumor-intrinsic YTHDF1 in driving immune evasion and its underlying mechanism.

No abstract

No abstract