Object-Attentional Untargeted Adversarial Attack
Chao Zhou, Yuan-Gen Wang, Guopu Zhu

TL;DR
This paper introduces an object-attentional untargeted adversarial attack that targets object regions for more imperceptible perturbations, improving attack efficiency and perceptual quality compared to existing methods.
Contribution
It proposes a novel object-attentional attack method combining object detection and saliency to enhance attack imperceptibility and efficiency.
Findings
Achieves better perceptual quality in adversarial examples.
Reduces query budget by up to 24.16%.
Effective on ImageNet-1K and COCO-Reduced-ImageNet datasets.
Abstract
Deep neural networks are facing severe threats from adversarial attacks. Most existing black-box attacks fool target model by generating either global perturbations or local patches. However, both global perturbations and local patches easily cause annoying visual artifacts in adversarial example. Compared with some smooth regions of an image, the object region generally has more edges and a more complex texture. Thus small perturbations on it will be more imperceptible. On the other hand, the object region is undoubtfully the decisive part of an image to classification tasks. Motivated by these two facts, we propose an object-attentional adversarial attack method for untargeted attack. Specifically, we first generate an object region by intersecting the object detection region from YOLOv4 with the salient object detection (SOD) region from HVPNet. Furthermore, we design an activation…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdversarial Robustness in Machine Learning · Anomaly Detection Techniques and Applications · COVID-19 diagnosis using AI
Methods+ ( 1 ) ⟷ 888 ⟷ ( 829 ) ⟷ 0881 How do I file a claim with Expedia? · BNB Customer Service Number +1-833-534-1729 · *Communicated@Fast*How Do I Communicate to Expedia? · Logistic Regression · Average Pooling · Feature Pyramid Network · Global Average Pooling · Softmax · 1x1 Convolution · k-Means Clustering
