4 August 2020 in IEEE Transactions on Multimedia
Turing test was originally proposed to examine whether machine?s behavior is indistinguishable from a human. The most popular and practical Turing test is CAPTCHA, which is to discriminate algorithm from human by offering recognitionalike questions. The recent development of deep learning has signi?cantly advanced the capability of algorithm in solving CAPTCHA questions, forcing CAPTCHA designers to increase question complexity. Instead of designing questions dif?cult for both algorithm and human, this study attempts to employ the limitations of algorithm to design robust CAPTCHA questions easily solvable to human. Speci?cally, our data analysis observes that human and algorithm demonstrates different vulnerability to visual distortions: adversarial perturbation is signi?cantly annoying to algorithm yet friendly to human. We are motivated to employ adversarially perturbed images for robust CAPTCHA design in the context of character-based questions. Four modules of multi-target attack, ensemble adversarial training, image preprocessing differentiable approximation, and expectation are proposed to address the characteristics of character-based CAPTCHA cracking. Qualitative and quantitative experimental results demonstrate the effectiveness of the proposed solution. We hope this study can lead to the discussions around adversarial attack/defense in CAPTCHA design and also inspire the future attempts in employing algorithm limitation for practical usage.
27 February 2023 in ACM Transactions on Multimedia Computing, Communications, and Applications
This study provides a new understanding of the adversarial attack problem by examining the correlation between adversarial attack and visual attention change. In particular, we observed that: (1) images with incomplete attention regions are more vulnerable to adversarial attacks; and (2) successful adversarial attacks lead to deviated and scattered activation map. Therefore, we use the mask method to design an attention-preserving loss and a contrast method to design a loss that makes the model’s attention rectification. Accordingly, an attention-based adversarial defense framework is designed, under which better adversarial training or stronger adversarial attacks can be performed through the above constraints. We hope the attention-related data analysis and defense solution in this study will shed some light on the mechanism behind the adversarial attack and also facilitate future adversarial defense/attack model design.
21 August 2023 in Information Sciences
26 June 2023 in Proceedings of the AAAI Conference on Artificial Intelligence
ImageNet pre-training has enabled state-of-the-art results on many tasks. In spite of its recognized contribution to generalization, we observed in this study that ImageNet pre-training also transfers adversarial non-robustness from pre-trained model into fine-tuned model in the downstream classification tasks. We first conducted experiments on various datasets and network backbones to uncover the adversarial non-robustness in fine-tuned model. Further analysis was conducted on examining the learned knowledge of fine-tuned model and standard model, and revealed that the reason leading to the non-robustness is the non-robust features transferred from ImageNet pre-trained model. Finally, we analyzed the preference for feature learning of the pre-trained model, explored the factors influencing robustness, and introduced a simple robust ImageNet pre-training solution. Our code is available at https://github.com/jiamingzhang94/ImageNet-Pretraining-transfers-non-robustness.
15 August 2025 in IEEE Transactions on Information Forensics and Security
The vulnerability of Deep Neural Networks (DNNs) to adversarial attacks poses a significant challenge to their deployment in safety-critical applications. While extensive research has addressed various attack scenarios, the no-box attack setting—where adversaries have no prior knowledge, including access to training data of the target model—remains relatively underexplored despite its practical relevance. This work presents a systematic investigation into leveraging large-scale Vision-Language Models (VLMs), particularly CLIP, as surrogate models for executing no-box attacks. Our theoretical and empirical analyses reveal a key limitation in the execution of no-box attacks stemming from insufficient discriminative capabilities for direct application of vanilla CLIP as a surrogate model. To address this limitation, we propose MF-CLIP (Margin-based Fine-tuned CLIP), a novel framework that enhances CLIP’s effectiveness as a surrogate model through margin-aware feature space optimization. Comprehensive evaluations across diverse architectures and datasets demonstrate that MF-CLIP substantially advances the state-of-the-art in no-box attacks, surpassing existing baselines by 15.23% on standard models and achieving a 9.52% improvement on adversarially trained models. Our code is made publicly available to facilitate reproducibility and future research in this direction.