CaptionFool: Universal Image Captioning Model Attacks
Image captioning models are encoder-decoder architectures trained on large-scale image-text datasets, making them susceptible to adversarial attacks. We present CaptionFool, a novel universal (input-agnostic) adversarial attack against state-of-the-art transformer-based captioning models. By modifying only 7 out of 577 image patches (approximately 1.2% of the image), our attack achieves 94-96% success rate in generating arbitrary target captions, including offensive content. We further demonstrate that CaptionFool can generate "slang" terms specifically designed to evade existing content moderation filters. Our findings expose critical vulnerabilities in deployed vision-language models and underscore the urgent need for robust defenses against such attacks. Warning: This paper contains model outputs which are offensive in nature.
Code (0)
등록된 구현이 없습니다.
Tasks
Adversarial AttackImage CaptioningSimilar Papers 제목 키워드 기반
Unveiling the Fragility of Vision-Language Models: Multi-Modal Adversarial Synergy via Texture-Constrained Perturbations and Cross-Modal Optimization
Large Vision-Language Models (LVLMs) have transformed multi-modal understanding, excelling in tasks like image captioning and visual question answering by integrating visual and textual inputs. However, their robustness …
Visual Question AnsweringAutonomous DrivingImage CaptioningControlled Caption Generation for Images Through Adversarial Attacks
Deep learning is found to be vulnerable to adversarial examples. However, its adversarial susceptibility in image caption generation is under-explored. We study adversarial examples for vision and language models, which …
Caption GenerationImage CaptioningLanguage ModellingProtect, Show, Attend and Tell: Empowering Image Captioning Models with Ownership Protection
By and large, existing Intellectual Property (IP) protection on deep neural networks typically i) focus on image classification task only, and ii) follow a standard digital watermarking framework that was conventionally …
Image Captioningimage-classificationImage ClassificationGeneralizing Universal Adversarial Attacks Beyond Additive Perturbations
The previous study has shown that universal adversarial attacks can fool deep neural networks over a large set of input images with a single human-invisible perturbation. However, current methods for universal adversaria…
Adversarial AttackExact Adversarial Attack to Image Captioning via Structured Output Learning with Latent Variables
In this work, we study the robustness of a CNN+RNN based image captioning system being subjected to adversarial noises. We propose to fool an image captioning system to generate some targeted partial captions for an imag…
Adversarial AttackImage Captioning