Deep neural networks (DNNs) have achieved remarkable performance in various tasks, including image classification. However, recent research has revealed the susceptibility of trained DNNs to subtle perturbations introduced into input images. Addressing these vulnerabilities is pivotal, leading to a significant area of study focused on developing attack algorithms capable of generating potent adversarial images. In scenarios where access to gradient information is restricted (black-box scenario), many existing methods introduce optimized perturbations to each individual pixels of an image to cause trained DNNs to mis-classify. However, due to the high-dimensional nature of this approach, current methods have inherent limitations. In contrast, our proposed approach involves the construction of perturbations by concatenating a series of overlapping semi-transparent shapes. Through the optimization of these shapes’ characteristics, we generate perturbations that result in the desired misclassification by the DNN. By conducting a series of attacks on state-of-the-art DNNs trained of CIFAR-10 and Imagenet datasets, our method consistently outperforms existing attack algorithms in terms of both query efficiency and success rate.

The use of evolutionary strategies in generating images is a common practice within the computational art community. In particular, a popular approach is to approximate a target image using overlapping, semi-transparent shapes and optimizing their attributes to increase similarity to the target image. However, existing methods usually require millions of fitness evaluations to construct good approximations. Within the evolutionary computation and machine learning communities, the use of surrogates has shown to decrease the number of fitness evaluations while achieving state-of-the-art results. Despite the gained traction of surrogate-assisted algorithms, their use within the computational art community is nonexistent. To address this, we extend the previous work of Bayesian Optimization by density-ratio estimation (BORE) to the image approximation task. By estimating the probability of improvement acquisition function using a convolutional probabilistic classifier, we search for solutions that maximize the acquisition function using an evolutionary strategy. By conducting experiments on six different styled target images, we demonstrate the superior performance achieved with the use of surrogate assistance.

Deep neural networks (DNNs) are susceptible to adversarial images, raising concerns about their reliability in safety-critical tasks. Sparse adversarial attacks, which limit the number of modified pixels, have shown to be highly effective in causing DNNs to misclassify. However, existing methods often struggle to simultaneously minimize the number of modified pixels and the size of the modifications, often requiring a large number of queries and assuming unrestricted access to the targeted DNN. In contrast, other methods that limit the number of modified pixels often permit unbounded modifications, making them easily detectable. To address these limitations, we propose a novel multi-objective sparse attack algorithm that efficiently minimizes the number of modified pixels and their size during the attack process. Our algorithm draws inspiration from evolutionary computation and incorporates a mechanism for prioritizing objectives that aligns with an attacker's goals. Our approach outperforms existing sparse attacks on CIFAR-10 and ImageNet trained DNN classifiers while requiring only a small query budget, attaining competitive attack success rates while perturbing fewer pixels. Overall, our proposed attack algorithm provides a solution to the limitations of current sparse attack methods by jointly minimizing the number of modified pixels and their size. Our results demonstrate the effectiveness of our approach in restricted scenarios, highlighting its potential to enhance DNN security.

Deep neural networks (DNNs) have demonstrated vulnerabilities to adversarial examples, which raises concerns about their reliability in safety-critical applications. While the majority of existing methods generate adversarial examples by making small modifications to the entire image, recent research has proposed a practical alternative known as adversarial patches. Adversarial patches have shown to be highly effective in causing DNNs to misclassify by distorting a localized area (patch) of the image. However, existing methods often produce clearly visible distortions since they do not consider the visibility of the patch. To address this, we propose a novel method for constructing adversarial patches that approximates the appearance of the area it covers. We achieve this by using a set of semi-transparent, RGB-valued circles, drawing inspiration from the computational art community. We utilize an evolutionary strategy to optimize the properties of each shape, and employ a simulated annealing approach to optimize the patch's location. Our approach achieves better or comparable performance to state-of-the-art methods on ImageNet DNN classifiers while achieving a lower $l_2$ distance from the original image. By minimizing the visibility of the patch, this work further highlights the vulnerabilities of DNNs to adversarial patches.

Evolutionary Algorithms (EA) are known to suffer from the curse of dimensionality resulting in poor performances when handling large-scale problems. Cooperative coevolution aims to overcome these issues in a divide and conquer approach by decomposing the original problem into several lower-dimensional sub-problems. For each sub-problem a chosen EA is applied for a defined number of function evaluations. This is repeated in a round robin like fashion until a terminating condition is met. A recently proposed area in the evolutionary computation field is the evolutionary multitask optimization (EMTO) framework. By jointly optimising several tasks, EMTO aims to exploit beneficial information across multiple tasks to improve the performance compared to optimising each task in isolation. In this paper, we consider a large-scale problem as a multi-task optimization problem by considering each sub-problem as an independent task. Applying an EMTO algorithm, knowledge transfer across sub-problems is carried out explicitly to improve the optimization of each sub-problem. We evaluate the effectiveness of our proposed algorithms empirically on a suite of separable and non-separable benchmark problems of varying dimensions.