Search Results (84)

Reliable scene understanding under open-world conditions requires intelligent perception systems to accurately recognize known semantic categories while remaining robust to out-of-distribution (OOD) inputs, distribution shifts, and uncertain environmental conditions. This challenge becomes increasingly important for resource-constrained edge intelligence, where lightweight models are expected to provide reliable predictions without sacrificing computational efficiency. Although knowledge distillation has achieved remarkable success in compressing deep neural networks, existing methods primarily transfer classification semantics and often neglect the uncertainty representations that are critical for reliable open-set perception. To address this issue, we propose Residual Feature-driven Knowledge Distillation (RFKD), a lightweight uncertainty-aware distillation framework for reliable open-set scene understanding under distribution shift. Instead of directly distilling output confidence or energy scores, RFKD reconstructs uncertainty within the student’s latent feature space through a compact residual uncertainty branch. The proposed framework combines confidence-aware supervision, relational uncertainty distillation, and energy-guided relative ordering to preserve teacher-induced uncertainty geometry while enabling the student to learn discriminative feature-level uncertainty representations. The present study is evaluated on unimodal image data and does not claim empirical validation for multimodal perception. Extensive experiments on CIFAR-100 using multiple OOD benchmarks demonstrate that RFKD consistently improves uncertainty estimation while maintaining high computational efficiency. Compared with the ResNet-50 Teacher (Energy), RFKD increases the average AUROC from 0.7793 to 0.8446 while reducing the model size from 23.71 M to 11.29 M parameters and computational complexity from 1.31 G to 0.56 G FLOPs; the ImageNet-style student baseline obtains an AUROC of 0.7528. A score-specific sensitivity analysis shows that the energy detector is strongest when the auxiliary ordering score uses the residual branch alone, whereas moderate coupling with the student’s log-sum-exp potential improves the standalone OOD head. These results demonstrate that explicitly modeling representation-level uncertainty offers an effective and efficient solution for reliable scene understanding, providing a practical reliability enhancement for future intelligent perception systems operating in open and dynamic environments. Full article

Biomimetic optimization transfers biological information-processing mechanisms into computational systems. Ant colony optimization (ACO) is a canonical example: artificial ants functionally abstract pheromone-mediated stigmergy, decentralized exploration, trail decay through algorithmic evaporation, and adaptive path reinforcement. Building on this functional biological analogue, we present a controlled cross-paradigm evaluation of routing-heuristic generation. A standardized interface embeds human-designed rules, the genetic programming hyper-heuristic GHPP, a resource-constrained DeepACO-MLP proxy, and an offline ReEvo-style proxy into the same ACO solver. The methods are evaluated on held-out TSP and CVRP instances in terms of solution quality, reported generation or training cost, interpretability, and cross-scale behavior under a matched distribution. GHPP yields the shortest routes at all tested scales; the ReEvo-offline proxy and strong human-designed rules generally form a second tier, whereas the resource-constrained neural proxy degrades markedly as problem size increases. These results do not establish an intrinsic ranking of full-capability paradigms. Instead, they show that method selection depends on the operating constraint and on evidence provenance: longer locally measured offline search favors GHPP, while auditable explicit rules characterize the human and ReEvo-offline proxies. By holding the ant-inspired execution mechanism fixed and varying the source of heuristic information, the benchmark clarifies how evolutionary, neural, and LLM-style design strategies interact with a common biomimetic substrate. Full article

Inverse kinematics (IK) is fundamental to robot motion planning. Classical analytical solvers require complete Denavit–Hartenberg (DH) parameters that are often proprietary or degraded by mechanical wear, and numerical solvers based on damped least squares (DLS) are sensitive to initialization, particularly near singularities. We propose XGNN, an explainable graph neural network positioned as a model-free, interpretable warm-start initializer for downstream numerical IK refinement rather than as a standalone replacement for analytical solvers. Each IK query is encoded as a 12-node graph in which six pose nodes and six joint nodes are connected through bipartite pose-to-joint attention edges and chain edges along the kinematic structure. GATv2 message passing aggregates information at each joint node; two ablation-validated design contributions (a learnable node-type embedding and an angle-aware composite loss) enable training to convergence. Evaluated on 300,000 trajectory-style samples generated from the ABB IRB 2400 kinematic model, XGNN achieves ${3.66}^{\circ }$ joint mean absolute error (MAE), comparable to a multilayer perceptron baseline (${3.09}^{\circ }$) and a bidirectional LSTM (${3.14}^{\circ }$) under identical training. The standalone joint accuracy of all learned models is too coarse for direct industrial use, but XGNN provides the strongest warm start for DLS refinement: the convergence rate improves from $98.4\%$ to $100\%$, mean iterations drop from 14.6 to 3.2, and wall-clock time per pose drops $5.0\times$ on the IRB 2400. The benefit transfers cross-platform to the Universal Robots UR5 collaborative manipulator (convergence rate $82.2\%$ to $100\%$, $10.0\times$ speedup) and survives DH parameter perturbation of up to $\pm 10\%$, simulating calibration drift or mechanical wear. The GATv2 attention coefficients additionally provide an interpretability signal at zero inference cost. XGNN therefore complements analytical and numerical IK methods as an interpretable, calibration-robust warm start when DH parameters are unavailable, proprietary, or degraded. Full article

Style transfer involves applying features from a stylized image to a content image, which has proven useful for image coloring. Significant progress has been made in utilizing neural networks for unsupervised sketch coloring using style transfer; however, existing models typically require user guidance. In this paper, we propose a GAN-Enhanced Style Transfer Network for Anime Sketch Colorization (STNGAN) that can fully automate coloring without user supervision. STNGAN incorporates a self-attention mechanism that enables the generator to capture global details and enhance color saturation and richness. The discriminator employs DenseNet to strengthen feature propagation and improve training stability. Additionally, we introduced a sketch reconstruction loss function to mitigate coloring overflow. Edge extraction was applied to obtain quantitative metrics. By contrasting light and dark areas between different color blocks and comparing them with the original sketches, we could objectively evaluate experimental performance using the peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) metrics. Ablation experiments were conducted to assess the impact of self-attention and DenseNet on coloring. The results indicate that the proposed method achieved consistent improvements over the selected baselines under the evaluated anime sketch colorization setting. Quantitative experiments showed that STNGAN achieved a PSNR of 10.560, an SSIM of 0.768, and an inception score (IS) of 1.755. Compared with the strongest competing method, pix2pixHD, STNGAN improved the PSNR by 6.8%, the SSIM by 1.7%, and the IS by 8.0%. A user study with 50 participants further confirmed the perceptual advantage of STNGAN, which obtained the highest mean opinion score (MOS) of 4.129. Full article

The closed-loop testing of autonomous driving systems critically depends on large-scale libraries of diverse and realistic 3D vehicle assets, yet current pipelines still rely on labor-intensive modeling or multi-view capture, making efficient construction a key bottleneck. To overcome this bottleneck and enable convenient, cost-effective 3D asset generation, we propose a semantic prior-guided framework for accurate and robust vehicle point cloud reconstruction from casually captured single-view photographs. Our framework is built on a diffusion backbone but is fundamentally driven by two forms of prior knowledge: First, geometric and appearance priors from camera-aware image features, masks, and distance-transform maps are projected onto the evolving point cloud, compensating for the severe information loss in single-view inputs. Second, we introduce distillation-style regulators—pretrained neural networks that encode vehicle type and model semantics; they act as teacher networks that impose high-level constraints on the generated point clouds, transferring rich semantic knowledge and effectively regularizing the learning process. With these priors, our model infers vehicle-specific semantics from limited observations and reconstructs high-quality 3D point cloud assets. On the 3DRealCar++ dataset, our method clearly surpasses state-of-the-art point cloud baselines in both F-score and Chamfer Distance. Full article

The purpose of this review is to present an overview of artificial intelligence methods for classifying paintings into the artistic movement to which they belong. To achieve this goal, a literature review of research articles from the 2014–2024 period was carried out. The search for scientific articles was carried out in the Scopus database. The initial search yielded 492 publications and after successive stages of screening and full-text evaluation, 39 articles were finally selected for detailed analysis. The review presents (a) the datasets used in the works, (b) the range of artistic movements examined and (c) the computational methods from machine learning to deep neural networks and transfer learning. Methodological issues are highlighted, such as class imbalance of the samples, dataset bias and the limitations of commonly used evaluation metrics. The general finding is that a variety of methodologies were applied, with an increasing use of deep learning and transfer learning models, which in many cases are reported as effective within specific datasets and experimental protocols. Finally, the review offers a taxonomy of methodologies and maps trends and research gaps in research on painting style classification over the last decade, while at the same time making suggestions for future research. Full article

Most artistic expressions are conveyed through images (e.g., painting) and audio (e.g., music), and deep learning has been successfully applied to neural style transfer within each of these modalities. However, there is still a lack of deep models that explicitly learn to transfer style between images and audio. Motivated by synesthesia, which reflects intrinsic connections between vision and hearing in the human brain, we propose a deep hybrid synesthesia model for audio–image style transfer. Our framework consists of two main components: (1) a component conversion module that learns cross-modal mappings between audio rhythm/spectrum and image color/shape in a continuous valence–arousal (VA) emotion space; and (2) a style conversion module that transfers high-level artistic styles between Eastern (ink-wash, shui-mo) and Western painting and their corresponding musical counterparts. We first learn emotion-aware feature networks that align low-level audio and visual components based on shared affective representations, and then model long-term stylistic structures for cross-modal style transfer. Experiments include “seeing the sound” (audio-to-image generation with controllable components) and full audio–image style transformations. Both objective analyses and subjective evaluations suggest that our model can produce cross-modal artworks whose perceived style and emotional content are consistent with human synesthetic impressions. 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

With the recent advancement of virtual try-on (VTO) technology, it is being applied to various fields. Although advancements in VTO technology have enabled not only 2D but also 3D visualization, applying style transfer to hyper-local regions remains challenging due to the complex surface curvature and ambiguous boundaries of 3D objects. To address these challenges, we propose an immersive 3D style transfer framework based on 3D Gaussian splatting. A segmentation model is employed to accurately segment target regions, and a large-scale specialized dataset is constructed to capture the morphological diversity of human hands. Furthermore, neural style transfer is integrated with the 3D representation to enable precise style application to hyper-local regions. The proposed framework achieves a mean intersection of union (mIoU) of 0.806 in segmentation and high-fidelity stylization with learned perceptual image patch similarity (LPIPS) and reference-based LPIPS (Ref-LPIPS) scores of 0.1472 and 0.0196, respectively. These results indicate that the proposed framework can provide the quality requirements and immersive VTO experience. Full article

Existing style-transfer steganography schemes suffer from three critical limitations: insufficient robustness against online social network (OSN) processing pipelines, susceptibility to steganalytic detection, and degraded visual quality. To address these challenges holistically, we propose StegTransfer—a unified framework that integrates: (1) forward non-differentiable distortion simulation, which emulates realistic OSN operations to enhance robustness; (2) adversarially hardened embedding through joint training with steganalyzers to improve security; and (3) payload-preserving style enhancement that optimizes visual aesthetics without sacrificing embedding capacity. Experimental evaluations demonstrate that StegTransfer achieves superior performance in visual fidelity (NIMA score: 6.32), robustness (PSNR up to 30.2 dB under JPEG compression), and security (detection rates as low as 15.5% and 62.3% under StegExpose and SiaStegNet, respectively. Full article

Wildfire smoke visualization using geostationary satellite imagery is essential for real-time monitoring and atmospheric analysis; however, inconsistencies in color tone across Geostationary Environment Monitoring Spectrometer (GEMS) images hinder reliable interpretation and model training. This study proposes a Standardized False Color Composite (SFCC) framework based on deep learning style transfer to enhance the visual consistency and interpretability of wildfire smoke scenes. Four tone-standardization methods were compared: the statistical Empirical Cumulative Distribution Function (ECDF) correction and three neural approaches—ReHistoGAN, StyTr², and Style Injection Diffusion Model (SI-DM). Each model was evaluated visually and quantitatively using six metrics (SSIM, LPIPS, FID, histogram similarity, ArtFID, and LSCI) and validated on three major wildfire events in Korea (2022–2025). Among the tested models, SI-DM achieved the most balanced performance, preserving structural features while ensuring consistent color-tone alignment (ArtFID = 1.620; LSCI mean = 0.894). Qualitative assessments further confirmed that SI-DM effectively delineated smoke boundaries and maintained natural background tones under complex atmospheric conditions. Additional analysis using GEMS UVAI, VISAI, and CHOCHO demonstrated that the styled composites partially reflect the optical and chemical characteristics distinguishing wildfire smoke from dust aerosols. The proposed SFCC framework establishes a foundation for visually standardized satellite smoke imagery and provides potential for future aerosol-type classification and automated detection applications. Full article

Within the traditional electronic neural network framework, Generative Adversarial Networks (GANs) have achieved extensive applications across multiple domains, including image synthesis, style transfer and data augmentation. Recently, several studies have explored the use of optical neural networks represented by the diffractive deep neural network (D²NN) for GANs. However, most of these focus on applications of the generative network, and there is currently no well-established D²NN architecture that simultaneously implements generative adversarial functionality. Here, we propose a novel implementation scheme for generative adversarial networks based on all-optical diffraction layers, demonstrating a complete all-optical adversarial architecture that simultaneously realizes both the generative network and the adversarial network (D²NN-GAN). We validated this method on the MNIST handwritten digit dataset, achieving Nash equilibrium convergence with the discriminator accuracy stabilizing around 50%. Concurrently, the average SSIM parameter of generated images reached 0.9573, indicating that the generated samples possess high quality and closely resemble real samples. Furthermore, we extended the framework to the KTH human action dataset, successfully reconstructing the “running” action with a discriminator accuracy of approximately 75%. The D²NN-GAN architecture introduces a fully optical generative adversarial model, providing a practical path for future optical modeling methods, such as image generation and video synthesis. Full article

Increasingly frequent intense rainfall is increasing landslide occurrence and risk. In southern China in particular, steep slopes and thin residual soils produce frequent landslide events with pronounced spatial heterogeneity. Therefore, displacement prediction methods that function across sites and deformation regimes in similar settings are essential for early warning. Most existing approaches adopt a multistage pipeline that decomposes, predicts, and recombines, often leading to complex architectures with weak cross-domain transfer and limited adaptability. To address these limitations, we present CRAFormer, a causal role-aware Transformer guided by a dynamic-lag Bayesian network-style causal graph learned from historical observations. In our system, the discovered directed acyclic graph (DAG) partitions drivers into five causal roles and induces role-specific, non-anticipative masks for lightweight branch encoders, while a context-aware Top-2 gate sparsely fuses the branch outputs, yielding sample-wise attributions. To safely exploit exogenous rainfall forecasts, next-day rainfall is entered exclusively through an ICS tail with a leakage-free block mask, a non-negative readout, and a rainfall monotonicity regularizer. In this study, we curate two long-term GNSS datasets from Guangxi (LaMenTun and BaYiTun) that capture slow creep and step-like motions during extreme rainfall. Under identical inputs and a unified protocol, CRAFormer reduces the MAE and RMSE by 59–79% across stations relative to the strongest baseline, and it lowers magnitude errors near turning points and step events, demonstrating robust performance for two contrasting landslides within a shared regional setting. Ablations confirm the contributions of the DBN-style causal masks, the leakage-free ICS tail, and the monotonicity prior. These results highlight a practical path from causal discovery to forecast-compatible neural predictors for rainfall-induced landslides. Full article

Ground-penetrating radar (GPR) is an important geophysical technique in subsurface detection. However, traditional numerical simulation methods such as finite-difference time-domain (FDTD) face challenges in accurately simulating complex heterogeneous mediums in real-world scenarios due to the difficulty of obtaining precise medium distribution information and high computational costs. Meanwhile, deep learning methods require excessive prior information, which limits their application. To address these issues, this paper proposes a novel two-step forward modeling strategy for GPR data of metal pipes. The first step employs the proposed Polarization Self-Attention Image Translation network (PSA-ITnet) for image translation, which is inspired by the process where a neural network model “understands” image content and “rewrites” it according to specified rules. It converts scene layout images (cross-sectional schematics depicting geometric details such as the size and spatial distribution of underground buried metal pipes and their surrounding medium) into simulated clutter-free GPR B-scan images. By integrating the polarized self-attention (PSA) mechanism into the Unet generator, PSA-ITnet can capture long-range dependencies, enhancing its understanding of the longitudinal time-delay property in GPR B-scan images. which is crucial for accurately generating hyperbolic signatures of metal pipes in simulated data. The second step uses the Polarization Self-Attention Style Transfer network (PSA-STnet) for style transfer, which transforms the simulated clutter-free images into data matching the distribution and characteristics of a real-world underground heterogeneous medium under unsupervised conditions while retaining target information. This step bridges the gap between ideal simulations and actual GPR data. Simulation experiments confirm that PSA-ITnet outperforms traditional methods in image translation, and PSA-STnet shows superiority in style transfer. Real-world experiments in a complex bridge support structure scenario further verify the method’s practicability and robustness. Compared to FDTD, the proposed strategy is capable of generating GPR data matching real-world subsurface heterogeneous medium distributions from scene layout models, significantly reducing time costs and providing an efficient solution for GPR data simulation and analysis. Full article

Deep learning has significantly advanced medical image analysis by enabling accurate, automated diagnosis across diverse clinical tasks such as lesion classification and disease detection. However, the practical deployment of these systems is still hindered by two major challenges: the limited availability of expert-annotated data and substantial domain shifts caused by variations in imaging devices, acquisition protocols, and patient populations. Although recent semi-supervised domain generalization (SSDG) approaches attempt to address these challenges, they often suffer from two key limitations: (i) reliance on computationally expensive uncertainty modeling techniques such as Monte Carlo dropout, and (ii) inflexible shared-head classifiers that fail to capture domain-specific variability across heterogeneous imaging styles. To overcome these limitations, we propose MultiStyle-SSDG, a unified semi-supervised domain generalization framework designed to improve model generalization in low-label scenarios. Our method introduces a multi-style ensemble pseudo-labeling strategy guided by entropy-based filtering, incorporates prototype-based conformity and semantic alignment to regularize the feature space, and employs a domain-specific multi-head classifier fused through attention-weighted prediction. Additionally, we introduce a dual-level neural-style transfer pipeline that simulates realistic domain shifts while preserving diagnostic semantics. We validated our framework on the ISIC2019 skin lesion classification benchmark using 5% and 10% labeled data. MultiStyle-SSDG consistently outperformed recent state-of-the-art methods such as FixMatch, StyleMatch, and UPLM, achieving statistically significant improvements in classification accuracy under simulated domain shifts including style, background, and corruption. Specifically, our method achieved 78.6% accuracy with 5% labeled data and 80.3% with 10% labeled data on ISIC2019, surpassing FixMatch by 4.9–5.3 percentage points and UPLM by 2.1–2.4 points. Ablation studies further confirmed the individual contributions of each component, and t-SNE visualizations illustrate enhanced intra-class compactness and cross-domain feature consistency. These results demonstrate that our style-aware, modular framework offers a robust and scalable solution for generalizable computer-aided diagnosis in real-world medical imaging settings. Full article