3260 papers • 126 benchmarks • 313 datasets
Adversarial attacks that are presented in the real world
(Image credit: Papersgraph)
These leaderboards are used to track progress in real-world-adversarial-attack-5
No benchmarks available.
Use these libraries to find real-world-adversarial-attack-5 models and implementations
No datasets available.
No subtasks available.
A novel easily reproducible technique to attack the best public Face ID system ArcFace in different shooting conditions by printing the rectangular paper sticker on a common color printer and putting it on the hat.
This study presents a realistic scenario in which an attacker influences algorithmic trading systems by using adversarial learning techniques to manipulate the input data stream in real time and presents various mitigation methods.
This study presents a framework that uses 3D modeling to craft adversarial patches for an existing real-world scene and evaluates its performance using a novel evaluation process that ensures that the results are reproducible in both the digital space and the real world.
This work shows that by playing the crafted adversarial perturbation as a separate source when the adversary is speaking, the practical speaker verification system will misjudge the adversary as a target speaker.
This paper proposes Segment and Complete defense (SAC), a general framework for defending object detectors against patch attacks through detection and removal of adversarial patches, and presents the APRICOT-Mask dataset, which augments the APRicOT dataset with pixel-level annotations of adversaria patches.
This work investigates two types of attacks -- goal hijacking and prompt leaking -- and demonstrates that even low-aptitude, but sufficiently ill-intentioned agents, can easily exploit GPT-3's stochastic nature, creating long-tail risks.
This article proposes a novel method to simultaneously optimize the position and perturbation for an adversarial patch, and thus obtain a high attack success rate in the black-box setting and extends this method to the traffic sign recognition task.
This work introduces flying adversarial patches, where multiple images are mounted on at least one other flying robot and therefore can be placed anywhere in the field of view of a victim multirotor and compares three methods that simultaneously optimize multiple adversarial patches and their position in the input image.
Adding a benchmark result helps the community track progress.