Adversarial training has proven to be effective in hardening networks against adversarial examples. However, the gained robustness is limited by network capacity and number of training samples. Consequently, to build more robust models, it is common practice to train on widened networks with more parameters. To boost robustness, we propose a conditional normalization module to adapt networks when conditioned on input samples. Our adaptive networks, once adversarially trained, can outperform their non-adaptive counterparts on both clean validation accuracy and robustness. Our method is objective agnostic and consistently improves both the conventional adversarial training objective and the TRADES objective. Our adaptive networks also outperform larger widened non-adaptive architectures that have 1.5 times more parameters. We further introduce several practical ``tricks'' in adversarial training to improve robustness and empirically verify their efficiency.
Transfer learning facilitates the training of task-specific classifiers using pre-trained models as feature extractors. We present a family of transferable adversarial attacks against such classifiers, generated without access to the classification head; we call these \emph{headless attacks}. We first demonstrate successful transfer attacks against a victim network using \textit{only} its feature extractor. This motivates the introduction of a label-blind adversarial attack. This transfer attack method does not require any information about the class-label space of the victim. Our attack lowers the accuracy of a ResNet18 trained on CIFAR10 by over 40\%.
Data poisoning--the process by which an attacker takes control of a model by making imperceptible changes to a subset of the training data--is an emerging threat in the context of neural networks. Existing attacks for data poisoning have relied on hand-crafted heuristics. Instead, we pose crafting poisons more generally as a bi-level optimization problem, where the inner level corresponds to training a network on a poisoned dataset and the outer level corresponds to updating those poisons to achieve a desired behavior on the trained model. We then propose MetaPoison, a first-order method to solve this optimization quickly. MetaPoison is effective: it outperforms previous clean-label poisoning methods by a large margin under the same setting. MetaPoison is robust: its poisons transfer to a variety of victims with unknown hyperparameters and architectures. MetaPoison is also general-purpose, working not only in fine-tuning scenarios, but also for end-to-end training from scratch with remarkable success, e.g. causing a target image to be misclassified 90% of the time via manipulating just 1% of the dataset. Additionally, MetaPoison can achieve arbitrary adversary goals not previously possible--like using poisons of one class to make a target image don the label of another arbitrarily chosen class. Finally, MetaPoison works in the real-world. We demonstrate successful data poisoning of models trained on Google Cloud AutoML Vision. Code and premade poisons are provided at https://github.com/wronnyhuang/metapoison
Meta-learning algorithms produce feature extractors which achieve state-of-the-art performance on few-shot classification. While the literature is rich with meta-learning methods, little is known about why the resulting feature extractors perform so well. We develop a better understanding of the underlying mechanics of meta-learning and the difference between models trained using meta-learning and models which are trained classically. In doing so, we develop several hypotheses for why meta-learned models perform better. In addition to visualizations, we design several regularizers inspired by our hypotheses which improve performance on few-shot classification.
To deflect adversarial attacks, a range of "certified" classifiers have been proposed. In addition to labeling an image, certified classifiers produce (when possible) a certificate guaranteeing that the input image is not an $\ell_p$-bounded adversarial example. We present a new attack that exploits not only the labelling function of a classifier, but also the certificate generator. The proposed method applies large perturbations that place images far from a class boundary while maintaining the imperceptibility property of adversarial examples. The proposed "Shadow Attack" causes certifiably robust networks to mislabel an image and simultaneously produce a "spoofed" certificate of robustness.
Adversarial patch attacks are among one of the most practical threat models against real-world computer vision systems. This paper studies certified and empirical defenses against patch attacks. We begin with a set of experiments showing that most existing defenses, which work by pre-processing input images to mitigate adversarial patches, are easily broken by simple white-box adversaries. Motivated by this finding, we propose the first certified defense against patch attacks, and propose faster methods for its training. Furthermore, we experiment with different patch shapes for testing, obtaining surprisingly good robustness transfer across shapes, and present preliminary results on certified defense against sparse attacks. Our complete implementation can be found on: https://github.com/Ping-C/certifiedpatchdefense.
Algorithmic trading systems are often completely automated, and deep learning is increasingly receiving attention in this domain. Nonetheless, little is known about the robustness properties of these models. We study valuation models for algorithmic trading from the perspective of adversarial machine learning. We introduce new attacks specific to this domain with size constraints that minimize attack costs. We further discuss how these attacks can be used as an analysis tool to study and evaluate the robustness properties of financial models. Finally, we investigate the feasibility of realistic adversarial attacks in which an adversarial trader fools automated trading systems into making inaccurate predictions.
Convex relaxations are effective for training and certifying neural networks against norm-bounded adversarial attacks, but they leave a large gap between certifiable and empirical robustness. In principle, convex relaxation can provide tight bounds if the solution to the relaxed problem is feasible for the original non-convex problem. We propose two regularizers that can be used to train neural networks that yield tighter convex relaxation bounds for robustness. In all of our experiments, the proposed regularizers result in higher certified accuracy than non-regularized baselines.