A White-Box False Positive Adversarial Attack Method on Contrastive Loss Based Offline Handwritten Signature Verification Models

arXiv:2308.08925v3 Announce Type: cross
Abstract: In this paper, we tackle the challenge of white-box false positive adversarial attacks on contrastive loss based offline handwritten signature verification models. We propose a novel attack method that treats the attack as a style transfer between closely related but distinct writing styles. To guide the generation of deceptive images, we introduce two new loss functions that enhance the attack success rate by perturbing the Euclidean distance between the embedding vectors of the original and …

adversarial adversarial attack adversarial attacks arxiv attack attacks box challenge cs.ai cs.cr cs.cv cs.lg false positive loss novel offline signature transfer verification