Search

We prove the first regularity theorem for the free boundary of solutions to shape optimization problems involving integral functionals, for which the energy of a domain Ω is obtained as the integral of a cost function j ( u ,  x ) depending on the solution u of a certain PDE problem on Ω . The main feature of these functionals is that the minimality of a domain Ω cannot be translated into a variational problem for a single (real or vector valued) state function. In this paper we focus on the case of affine cost functions j ( u , x ) = - g ( x ) u + Q ( x ) , where u is the solution of the PDE - Δ u = f with Dirichlet boundary conditions. We obtain the Lipschitz continuity and the non-degeneracy of the optimal u from the inwards/outwards optimality of Ω and then we use the stability of Ω with respect to variations with smooth vector fields in order to study the blow-up limits of the state function u . By performing a triple consecutive blow-up, we prove the existence of blow-up sequences converging to homogeneous stable solution of the one-phase Bernoulli problem and according to the blow-up limits, we decompose ∂ Ω into a singular and a regular part. In order to estimate the Hausdorff dimension of the singular set of ∂ Ω we give a new formulation of the notion of stability for the one-phase problem, which is preserved under blow-up limits and allows to develop a dimension reduction principle. Finally, by combining a higher order Boundary Harnack principle and a viscosity approach, we prove C ∞ regularity of the regular part of the free boundary when the data are smooth., Competing Interests: Conflict of interestThe authors have no conflict of interest to declare., (© The Author(s) 2024.)

The existence of real-world adversarial examples (RWAEs) (commonly in the form of patches) poses a serious threat for the use of deep learning models in safety-critical computer vision tasks such as visual perception in autonomous driving. This article presents an extensive evaluation of the robustness of semantic segmentation (SS) models when attacked with different types of adversarial patches, including digital, simulated, and physical ones. A novel loss function is proposed to improve the capabilities of attackers in inducing a misclassification of pixels. Also, a novel attack strategy is presented to improve the expectation over transformation (EOT) method for placing a patch in the scene. Finally, a state-of-the-art method for detecting adversarial patch is first extended to cope with SS models, then improved to obtain real-time performance, and eventually evaluated in real-world scenarios. Experimental results reveal that even though the adversarial effect is visible with both digital and real-world attacks, its impact is often spatially confined to areas of the image around the patch. This opens to further questions about the spatial robustness of real-time SS models.

Artificial intelligence is making extraordinary progress with an unprecedented rate, reaching and surpassing human capabilities in many tasks previously considered unattainable by machines, as language translation, music composition, object detection, medical diagnoses, software programming, and many others. Some people are excited about these results, while others are raising serious concerns for possible negative impacts in our society. This article addresses several questions that are often raised about intelligent machines: Will machines ever surpass human intellectual capacities? What will happen next? What will be the impact in our society? What are the jobs that artificial intelligence puts at risk? Reasoning about these questions is of fundamental importance to predict possible future scenarios and prepare ourselves to face the consequences., Competing Interests: The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2023 Buttazzo.)

Although Deep Neural Networks (DNNs) have shown incredible performance in perceptive and control tasks, several trustworthy issues are still open. One of the most discussed topics is the existence of adversarial perturbations, which has opened an interesting research line on provable techniques capable of quantifying the robustness of a given input. In this regard, the euclidean distance of the input from the classification boundary denotes a well-proved robustness assessment as the minimal affordable adversarial perturbation. Unfortunately, computing such a distance is highly complex due the non-convex nature of DNNs. Despite several methods have been proposed to address this issue, to the best of our knowledge, no provable results have been presented to estimate and bound the error committed. This paper addresses this issue by proposing two lightweight strategies to find the minimal adversarial perturbation. Differently from the state-of-the-art, the proposed approach allows formulating an error estimation theory of the approximate distance with respect to the theoretical one. Finally, a substantial set of experiments is reported to evaluate the performance of the algorithms and support the theoretical findings. The obtained results show that the proposed strategies approximate the theoretical distance for samples close to the classification boundary, leading to provable robustness guarantees against any adversarial attacks.

Over the past few years, convolutional neural networks (CNNs) have proved to reach superhuman performance in visual recognition tasks. However, CNNs can easily be fooled by adversarial examples (AEs), i.e., maliciously crafted images that force the networks to predict an incorrect output while being extremely similar to those for which a correct output is predicted. Regular AEs are not robust to input image transformations, which can then be used to detect whether an AE is presented to the network. Nevertheless, it is still possible to generate AEs that are robust to such transformations. This article extensively explores the detection of AEs via image transformations and proposes a novel methodology, called defense perturbation, to detect robust AEs with the same input transformations the AEs are robust to. Such a defense perturbation is shown to be an effective counter-measure to robust AEs. Furthermore, multinetwork AEs are introduced. This kind of AEs can be used to simultaneously fool multiple networks, which is critical in systems that use network redundancy, such as those based on architectures with majority voting over multiple CNNs. An extensive set of experiments based on state-of-the-art CNNs trained on the Imagenet dataset is finally reported.

We describe the development of a novel wrist-mounted apnea dive computer. The device is able to measure and display transcutaneous oxygen saturation, heart rate, plethysmographic pulse waveform, depth, time and temperature during breath-hold dives. All measurements are stored in an external memory chip. The data-processing software reads from the chip and writes the processed data into a comma-separated values file which can be analysed by applications such as Microsoft Excel™ or Open Office™. The housing is waterproof and pressure-resistant to more than 20 bar (2.026 MPa) (breath-hold divers have already exceeded 200 metres' sea water depth). It is compact, lightweight, has low power requirements and is easy to use.

If human consciousness is the result of complex neural electro-chemical interactions occurring in the brain, the question of whether a machine can ever become self-aware could be a matter of time: the time necessary to fully understand the functional behavior of the brain structure, develop a mathematical model of it, and implement an artificial system capable of working according to such a model. This paper addresses several issues related to the possibility of developing a conscious artificial brain. A number of hazardous questions are posed to the reader, each addressing a specific technical or philosophical issue, which is discussed and developed in a form of a hazardous answer.