Fooling Automated Surveillance Cameras: Adversarial Patches to Attack Person Detection

By Simen Thys, Wiebe Van Ranst, Toon Goedemé. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2019

737 citations recorded by the source corpus.

Adversarial attacks on machine learning models have seen increasing interest in the past years. By making only subtle changes to the input of a convolutional neural network, the output of the network can be swayed to output a completely different result. The first attacks did this by changing pixel values of an input image slightly to fool a classifier to output the wrong class. Other approaches have tried to learn "patches" that can be applied to an object to fool detectors and classifiers. Some of these approaches have also shown that these attacks are feasible in the real-world, i.e. by modifying an object and filming it with a video camera. However, all of these approaches target classes that contain almost no intra-class variety (e.g. stop signs). The known structure of the object is then used to generate an adversarial patch on top of it. In this paper, we present an approach to generate adversarial patches to targets with lots of intra-class variety, namely persons. The goal is to generate a patch that is able successfully hide a person from a person detector. An attack that could for instance be used maliciously to circumvent surveillance systems, intruders can sneak around undetected by holding a small cardboard plate in front of their body aimed towards the surveilance camera. From our results we can see that our system is able significantly lower the accuracy of a person detector. Our approach also functions well in real-life scenarios where the patch is filmed by a camera. To the best of our knowledge we are the first to attempt this kind of attack on targets with a high level of intra-class variety like persons.

Subfields: Artificial Intelligence, Signal Processing

View Paper: https://www.semanticscholar.org/paper/55c5e4753ac6bfa88d2f0e00c8e881c49b72d48e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Signal Processing

Citations in the corpus

AdvHat: Real-World Adversarial Attack on ArcFace Face ID System (2019, methodology): “of such works [35] proposes to use EOT, NPS, and TV loss for fooling YOLOv2-based person detector.”
Addressing Privacy Threats from Machine Learning (2021, background): “In recent years, researchers have developed a number of strategies for evading facial recognition using adversarial examples (Harvey, 2017; Thys et al., 2019; Cherepanova et al., 2021; Li and Choi, 2021; Chandrasekaran et al., 2021).”
Making an Invisibility Cloak: Real World Adversarial Attacks on Object Detectors (2019, background): “produce printed adversarial patches [31] that deceive person detectors instantiated by Yolov2 [26].”
Robust Feature-Level Adversaries are Interpretability Tools (2021, background): “More recently, [57, 6, 15, 2, 37, 67, 32, 31] created adversarial clothing, stickers, or objects.”
Adversarial Examples on Object Recognition (2020, neutral): “impliesdirectconsequences on security. Other tasks such as reinforcement learning [10, 80, 106], speech recognition [23, 27], facial recognition [150], semantic segmentation [178] or video processing [103, 164, 170] are also explored. Moreover, some practical experiments are not covered in detail, e”
Fawkes: Protecting Privacy against Unauthorized Deep Learning Models (2020, background): “Others propose “adversarial patches” that target “person identification” models, making it difficult for models to recognize the wearer as a person in an image [54, 64].”
A Brief History of Artificial Intelligence: On the Past, Present, and Future of Artificial Intelligence (2019, neutral): “24 While China and, to a certain extent, the United States try to limit the barriers for firms to use and explore AI, the European Union has taken the opposite direction with the introduction of the General Data Protection Regulation (GDPR) that significantly limits the way in which personal…”
Adversarial T-Shirt! Evading Person Detectors in a Physical World (2019, methodology): “n both digital and physical worlds and show that the proposed adversarial Tshirt achieves79% and 63%attack success rates respectively when attacking YOLOv2. By contrast,the physical adversarial patch [14] printed on a T-shirtonly achieves27%attack success rate. Some of our results are highlighted in”