Search Results (58)

Deep neural networks are vulnerable to adversarial examples, even in hard-label black-box settings where only the top-1 prediction is available. To address the challenges of high-dimensional optimization under limited query budgets, we propose two query-efficient attack methods: Adam-OPT, which integrates Adam-based adaptive optimization into the ray-search framework to stabilize and accelerate zeroth-order gradient updates; Prior-Adam-OPT, which further incorporates transfer-based priors from surrogate models to enhance gradient estimation. Adam-OPT leverages historical gradient information and per-parameter adaptive updates to improve convergence, while Prior-Adam-OPT constructs a prior-guided orthogonal search basis that combines surrogate and random directions, enhancing both gradient accuracy and query efficiency. Our approach demonstrates superior performance across CIFAR-10, ImageNet, and zero-shot CLIP models, consistently reducing perturbation magnitudes and improving attack efficiency compared to state-of-the-art hard-label attacks. Ablation studies highlight the importance of the number of vectors used for gradient estimation and the quality of surrogate models, showing that combining adaptive optimization with transfer-based priors provides a scalable and robust framework for generating high-quality adversarial examples in challenging black-box scenarios. Full article

The highly nonlinear aerodynamic interference in tandem-airfoil configurations significantly hinders the precise exploitation of their aerodynamic potential. To address this issue, this study establishes a high-fidelity computational fluid dynamics benchmark. A high-quality sample set is constructed using Latin hypercube sampling combined with an intra-layer replacement strategy. Subsequently, a Gaussian process surrogate model and Bayesian optimization are employed to maximize the total system lift coefficient across a four-dimensional design space comprising longitudinal and vertical separations, fore airfoil angle of attack, and angle of attack difference. Global sensitivity analysis indicates that longitudinal separation dominates the interference modes. Optimization reveals a distinct mode switching phenomenon using a longitudinal separation of twice the chord length as the critical threshold. In the close-coupled configuration, a negative optimal angle of attack difference enhances the slot effect and upwash induction, thereby delaying rear airfoil stall and achieving synergistic lift enhancement. Conversely, in the distant-coupled configuration, the system transitions to a decoupled compensation mode, where a positive angle of attack difference compensates for the effective angle of attack loss induced by wake downwash. This research elucidates the competitive mechanisms between inter-airfoil slot flow and wake interference, providing a theoretical reference for the aerodynamic layout optimization of tandem-airfoil aircraft. Full article

The aerodynamic performance of compressor stators critically affects aircraft engine efficiency, yet traditional CFD-based evaluation and optimization suffer from high computational cost. This study addresses this gap by developing deep learning surrogate models to predict total pressure loss coefficient and outlet flow angle deviation for compressor stator vanes, using two geometric parameters—stagger angle ${\beta }_y$, leading-edge radius ratio R_rle, and one operational parameter, attack angle α. A high-fidelity dataset of 1701 cases was generated via automated CFD simulations using the transitional SST k-ω model. Among evaluated models—including standard CNN, CBAM-CNN, SS-CNN, and CNN-Transformer, SS-CNN achieved the highest accuracy, reducing mean absolute percentage error from 3.56% to 2.03% for loss and from 1.49% to 1.11% for outlet angle, with substantial computational savings. These surrogate models were integrated into a multi-objective optimization framework. The optimized vane, featuring a reduced leading-edge radius ratio within a stable stagger range, reduced total pressure loss by 2.38% (from 0.0570 to 0.0556) at the design attack angle of −2.83°, while the outlet angle deviation decreased from 0.439° to 0.066° (85% reduction), with the outlet angle improvement concentrated near the design condition. This work demonstrates a systematic, data-driven pipeline combining parametric modeling, automated simulation, deep learning-based prediction, and rapid optimization, offering an efficient solution for intelligent compressor blade design. Full article

The remarkably developed graph neural networks (GNNs) are extensively applied to specific tasks in online social networks (OSNs), especially in the vital domain of social trust. Meanwhile, the vulnerability of GNN applied in trust assessment can be exposed leveraging the deployment of subtly designed adversarial attacks. However, the predominant adversarial attack strategies targeting GNN are manipulating graph structure, which is not well-suited for social trust prediction tasks. In this article, we craft a novel black-box attack strategy, T-Attack, aimed at trust evaluation tasks, without tampering with the network structure of the specific trust prediction models. Specifically, a surrogate model is initially established to replicate trust prediction models based on GNN. The attack strategy on the surrogate model is formulated by adding unnoticed perturbations to user features related to network structure and manipulating the existing trust rating based on prior knowledge of social trust propagation, thereby avoiding a traditional attack against the GNN-based trust prediction model via modifying graph structure. By leveraging transferable attacks, our attack strategy can also distort the predictions of GNN-based trust prediction models. Through implementing extensive experiments in untargeted attack scenarios, we demonstrate the predictive performance of our crafted surrogate model and verify the effectiveness of the attack strategy on various GNN-based trust prediction models. Full article

Automated mobile microscopy in Internet of Things (IoT) networks is essential for scaling malaria screening in resource-constrained environments. Deploying standard convolutional architectures here introduces severe adversarial vulnerabilities. Post-Training Quantization (PTQ) mitigates hardware constraints by converting floating-point models to 8-bit integers (INT8); however, its impact on clinical safety and security remains unexplored. This study presents an adversarial audit of quantized Vision Transformers for medical edge deployment. We evaluated a Swin-Tiny transformer against ViT-Tiny and MobileNetV3 baselines using a 27,558-image malaria dataset and an out-of-distribution (OOD) White Blood Cell dataset. Our findings redefine the “Quantization Shield” hypothesis. PTQ compresses the Swin model by 3.9× (to 27.89 MB) with a negligible 0.11% accuracy drop, maintaining statistical reliability on OOD tests. However, the hypothesized architectural resilience shatters under white-box Projected Gradient Descent (PGD) attacks. Despite robustness against single-step attacks, both MobileNetV3 and the INT8 Swin-Tiny collapse to 0.00% accuracy under iterative PGD. Conversely, the quantized Swin-Tiny resists black-box transfer attacks from a surrogate, maintaining 81.00% accuracy. We conclude that while quantized Vision Transformers meet mobile sensor constraints, integer quantization provides zero innate defense against targeted iterative perturbations, exposing a critical vulnerability in diagnostic IoT networks. Full article

DNNs are highly vulnerable to adversarial examples (AEs). To achieve high transferability, traditional AEs often introduce unnatural artifacts that are easily perceptible to the human eye. Unrestricted attacks have emerged as a promising paradigm to generate more natural unrestricted adversarial examples (UAEs). However, existing UAEs struggle to balance visual fidelity and black-box transferability. Color-based attacks produce noticeable unnatural visual mutations, and diffusion-based attacks transfer poorly to unknown black-box models. We observe that directly injecting unconstrained random perturbations into the diffusion latent space destroys the normal distribution of data, thereby causing a distribution shift. Distribution shifts degrade adversarial perturbations into invalid noise and cause surrogate model overfitting. Furthermore, introducing elastic deformation during the denoising process forces surrogate models to focus on highly transferable features. As a result, we propose an unrestricted attack based on deformation-constrained diffusion, called DeDiAttack. Our method utilizes the manifold prior knowledge of diffusion models to translate elastic deformations into smooth fluid changes. The mechanism effectively eliminates unnatural artifacts and generates highly natural and transferable UAEs. Extensive black-box experiments demonstrate that DeDiAttack outperforms existing attacks and improves the black-box transferability of generated UAEs by 7.2% on the ViT-B surrogate model. The proposed method also provides a useful robustness evaluation tool for vision-based sensing and imaging systems. Full article

Adversarial attacks have been extensively studied in recent years to investigate the vulnerability mechanisms of deep neural networks and enhance model robustness and security. However, the transferability of adversarial examples across different models remains a fundamental challenge in black-box attacks. Existing ensemble attack methods primarily aggregate gradient information from multiple surrogate models through simple averaging, failing to consider gradient conflicts and cancellations among heterogeneous models, which results in poor transferability. To address this limitation, we propose a novel ensemble adversarial attack method called Region-wise PCGrad and Margin-Guided Adaptive Weighting Ensemble Attack (RPMGEA). To tackle gradient conflicts, we adopt a region-wise PCGrad method that divides gradient maps into semantically relevant regional blocks for conflict resolution. To address weight allocation issues, we directly measure transferability contributions by evaluating decision boundary changes caused by temporary adversarial examples generated from each model’s gradients across all models, thereby adaptively allocating weights to the models. RPMGEA significantly enhances the transferability of ensemble attacks, achieving average attack success rates of up to 93.7% and 90.4% on conventional and adversarial training models, respectively, and up to 88.9% on defense models, demonstrating superior performance compared to existing state-of-the-art ensemble attack methods. Full article

Adversarial transferability makes black-box attacks practical and exposes weaknesses of deep neural networks for computer vision, image recognition, and visual understanding. Among various transferability-enhancing methods, input transformation is one of the most effective strategies. However, existing methods often ignore the decoupling of style and semantics in the input image, as well as the need for customized transformation strategies, resulting in limited performance gains or suboptimal outcomes. In this paper, we propose a novel Fourier-based perspective for input transformation generalization in the context of vision adversarial attacks. The main observations are that the Fourier amplitude captures stylistic information and the phase encompasses richer semantics which are crucial for visual understanding. Motivated by this, we develop a Fourier-based strategy, which performs a stylistic transform and semantic mixup on the input examples to improve transferability. To avoid inconsistent semantics of augmented images for the surrogate model, we mix the original images with the augmentations to maintain semantic consistency and mitigate imprecise gradients. Extensive experiments on ImageNet-compatible datasets demonstrate that our method consistently outperforms existing input transformation attacks. Full article

Adversarial examples reveal critical vulnerabilities in deep neural networks, posing significant risks in real-world deployment. In black-box settings, transferable targeted attacks rely on surrogate models but often suffer from low success rates. We argue that this limitation arises not only from surrogate-boundary overfitting but also from insufficient alignment with the target semantic space, which restricts the ability of adversarial examples to encode target-specific characteristics. To address this issue, we propose Cosine-Based Logit Alignment (CBLA), a unified framework for transferable targeted attacks. CBLA replaces the conventional cross-entropy loss with a cosine similarity objective to enhance directional alignment in logit space and alleviate gradient saturation. In addition, a semantic-invariant transformation strategy is introduced to improve structural consistency and cross-model generalization. Experiments on the ImageNet validation set demonstrate that CBLA consistently improves targeted attack success rates, achieving an average gain of 4.55% over strong baselines across multiple architectures. Full article

The development of sustainable and efficient aircraft concepts, such as those featuring flared folding wing tips (FWTs), introduces both aerodynamic and structural challenges. FWTs have demonstrated strong potential for enhancing aerodynamic performance and alleviating gust-induced loads, making them an attractive option for next-generation transport aircraft. This study investigates the load reduction potential of transonic transport aircraft configurations equipped with hinged FWTs, with particular focus on gust impact. Reynolds Averaged Navier Stokes simulations are combined with Gaussian Process regression to evaluate the influence of the fold angle, flare angle, and angle of attack on key quantities of interest, including lift and wing root bending moment coefficients. The GP surrogate model, developed within the Gust Load Alleviation by Non-linear Folding Wing Tip (GUSTAFO) project, accurately reproduces the high-fidelity data while capturing the underlying system uncertainties. The results show that increasing the flare angle within a given folding deflection can reduce the wing root bending moment by up to 38% for flare angles between 0–${45}^{\circ }$ and fold angles between 0–${15}^{\circ }$. These findings highlight the effectiveness of surrogate-based modeling for early-stage design and emphasize the importance of incorporating FWT behavior to achieve accurate, efficient, and robust load predictions. Full article

To address the strong nonlinearity of aerodynamic loads during projectile exterior ballistic flight and the difficulty in accurately modeling fuze dynamic responses, this paper proposes a data-driven modeling and simulation method for fuze exterior ballistic dynamics. A high-fidelity aerodynamic database covering a range of Mach numbers and angles of attack is constructed based on CFD (Computational Fluid Dynamics) simulations. An MLP (Multilayer Perceptron) neural network is then employed to develop an aerodynamic surrogate model, enabling continuous representation of aerodynamic loads within the given sample space. The results show that, within the data coverage range, the proposed model is able to capture the nonlinear variation in aerodynamic parameters and shows improved prediction accuracy compared with the polynomial fitting method. Specifically, for typical aerodynamic parameters, the RMSE (Root Mean Square Error) is reduced from 5.758 to 0.223, the MAE (Mean Absolute Error) is reduced to 0.099, and the R² (Coefficient of Determination) approaches 1. On this basis, the aerodynamic surrogate model is embedded into a six-degree-of-freedom projectile–fuze exterior ballistic dynamics model via the secondary development interface of ADAMS 2020 (Automated Dynamic Analysis of Mechanical Systems), enabling coupled simulation between aerodynamic loads and multibody dynamics. Comparison with firing table data indicates that, under typical operating conditions, the relative deviation of ballistic parameters is generally better than 94%, demonstrating that the proposed method can reasonably reproduce the projectile exterior ballistic characteristics. Furthermore, based on the coupled dynamics model, the dynamic response characteristics of the fuze moving body during the exterior ballistic phase are analyzed. The results indicate that the axial forward overload of the moving body increases significantly with the initial nutation angle, and the variation in the axial projection of gravity induced by nutation plays an important role in its transient response. The proposed approach provides a useful reference for the dynamic response analysis and safety evaluation of fuzes. Full article

Adversarial examples have revealed the vulnerability of deep neural networks, and their transferability makes black-box attacks particularly concerning. However, perturbations crafted on a surrogate model often do not remain sufficiently effective on unseen target models. In this paper, we revisit this issue from a frequency-domain perspective and observe that perturbation optimization can become overly dependent on specific spectral patterns, which weakens cross-model transfer. To address this problem, we propose frequency-consistent regularization (FCR), a simple plug-in strategy that can be combined with existing iterative attacks. FCR introduces multiple low-frequency preserving views with randomly sampled frequency ranges at each iteration and optimizes perturbations across these varied views. In this way, the generated perturbations are less tied to a specific frequency configuration and show improved transferability. Experimental results show that FCR consistently improves the transfer performance of various iterative attacks. The improvement is observed not only in standard target models but also in adversarially trained models, where the gain is often more pronounced. Full article

Numerical simulations employing the dynamic mesh method were performed to investigate the unsteady aerodynamics of variable-sweep wings during morphing. Quasi-steady and unsteady aerodynamic characteristics were compared, and the effects of key operating conditions (freestream velocity, angle of attack, morphing period, wingspan, chord length) on unsteady aerodynamics were analyzed. To enable the rapid prediction of unsteady aerodynamics, a Kriging surrogate model was established and validated against high-fidelity CFD results. The results indicate that unsteady effects manifest as hysteresis loops in aerodynamic coefficients within the morphing cycle. The wing morphing period, angle of attack, freestream velocity, and wingspan have a pronounced impact on the unsteady aerodynamic characteristics, whereas the effect of chord length is negligible. Reduced morphing periods, increased angles of attack, and increased wingspans amplify the hysteresis loop size and enhance the unsteady effects. An increase in the freestream velocity intensifies unsteady effects in the subsonic flow, while it attenuates unsteady effects in the supersonic flow. Compared to direct CFD simulations, the Kriging model for unsteady aerodynamic characteristics prediction achieves a 97% improvement in overall computational efficiency, while its predicted hysteresis loops are in good agreement with CFD results in both trend and magnitude, with an average prediction error below 4% and a maximum error of less than 6%. The Kriging surrogate model developed in this study offers substantial practical value for engineering applications by meeting the demand for rapid aerodynamic computation in the concept design phase for morphing aircraft. Full article

Fast and reliable aerodynamic predictions are crucial in the early phases of aircraft design, where a quick assessment of various configurations is required. In this context, the Vortex Lattice Method (VLM) is widely adopted due to its computational efficiency; however, its predictive accuracy is highly sensitive to panel discretization strategies, which are often determined heuristically. This study proposes a bio-inspired optimization framework for VLM panel discretization and evaluates it through a systematic case study on a representative wing geometry. A grid-convergence analysis was initially carried out to ensure solution independence across various spanwise-to-chordwise panel ratios. Subsequently, a novel Hybrid Response Surface Methodology (HRSM), integrating Box–Behnken and Central Composite experimental designs, was employed to enable a more comprehensive exploration of the factor space while quantifying the effects of clustering parameters at the leading-edge, trailing-edge, root, and tip regions of the wing. The HRSM dataset was further utilized to train Ensemble Machine-Learning surrogate models, which were coupled with bio-inspired Crayfish and Aquila optimization algorithms, alongside a classical Genetic Algorithm (GA) as a performance benchmark, to identify the optimal discretization strategy and to enable a comparative assessment of their convergence behavior and robustness against the numerical noise of the ensemble-based landscape. Compared to base (i.e., uniform) panel distribution, the optimally clustered discretization enhanced overall aerodynamic prediction accuracy by approximately 33%, particularly at low angles of attack, while maintaining robust performance at higher angles. Both algorithms converged to similar minima; however, the Aquila algorithm achieved higher solution consistency, whereas the Crayfish algorithm exhibited greater dispersion despite faster convergence, revealing a multimodal optimization landscape. The variance decomposition revealed that trailing-edge clustering dominated aerodynamic accuracy at low angles of attack, contributing up to 90% of the total variance, whereas tip clustering became increasingly influential at higher angles, exceeding 30%, highlighting the need for adaptive discretization strategies to ensure reliable VLM-based aerodynamic analyses. Full article

No-reference image quality assessment (NR-IQA) models achieve high correlation with human mean opinion scores (MOS) on clean benchmarks, yet recent work shows they can be highly vulnerable to small adversarial perturbations that severely degrade ranking consistency, including in black-box settings. We introduce the Spectral Robustness Mixer (SRM), a lightweight neck inserted between an NR-IQA backbone and regression head, designed to reduce adversarial sensitivity without changing the dataset, label format, or target metric. SRM couples (i) deep-to-shallow cross-scale fusion via a Nyström low-rank attention surrogate, (ii) ridge-conditioned landmark kernels with ridge regularization, solved via numerically stable small-matrix factorization (SVD/LU) to improve conditioning, and (iii) variance-aware entropy-regularized fusion gates with a bounded gain cap to limit gradient amplification. We evaluate SRM on TID2013 and KonIQ-10k under a white-box ${\mathcal{l}}_{\infty }/{\mathcal{l}}_2$ attack ensemble that includes per-image regression objectives and a correlation-aware pairwise inversion objective (a ranking-inspired surrogate for correlation inversion), with expectation-over-transformation (EOT) and anti-gradient masking checks. At $ϵ=4/255$ (${\mathcal{l}}_{\infty }$), SRM improves worst-case robust Spearman’s rank-order correlation coefficient (SROCC; defined as the minimum over our fixed attack ensemble) by an absolute $0.06$–$0.08$ SROCC points (i.e., correlation-coefficient units, not percentage gain) across datasets/backbones, while keeping clean SROCC within $0.00$–$0.01$ of the baseline. We observe similar trends for Pearson linear correlation coefficient (PLCC). Full article