Ask, Attend, Attack: A Effective Decision-Based Black-Box Targeted Attack for Image-to-Text Models

While image-to-text models have demonstrated significant advancements in various vision-language tasks, they remain susceptible to adversarial attacks. Existing white-box attacks on image-to-text models require access to the architecture, gradients, and parameters of the target model, resulting in low practicality. Although the recently proposed gray-box attacks have improved practicality, they suffer from semantic loss during the training process, which limits their targeted attack performance. To advance adversarial attacks of image-to-text models, this paper focuses on a challenging scenario: decision-based black-box targeted attacks where the attackers only have access to the final output text and aim to perform targeted attacks. Specifically, we formulate the decision-based black-box targeted attack as a large-scale optimization problem. To efficiently solve the optimization problem, a three-stage process textitAsk, Attend, Attack, called textitAAA, is proposed to coordinate with the solver. textitAsk guides attackers to create target texts that satisfy the specific semantics. textitAttend identifies the crucial regions of the image for attacking, thus reducing the search space for the subsequent textitAttack. textitAttack uses an evolutionary algorithm to attack the crucial regions, where the attacks are semantically related to the target texts of textitAsk, thus achieving targeted attacks without semantic loss. Experimental results on transformer-based and CNN+RNN-based image-to-text models confirmed the effectiveness of our proposed textitAAA.