Search Results (587)

In this paper, a security-enhanced three-dimensional (3D) image hiding and encryption method is proposed by combining a layer-based phase-only hologram (POH) under structured light illumination with chaotic encryption and digital image watermarking technology. In the proposed method, the original 3D plaintext image is firstly encoded into a layer-based POH and then further encrypted into an encrypted phase with the help of a chaotic random phase mask (CRPM). Subsequently, the encrypted phase is embedded into a visible ciphertext image by using a digital image watermarking technology based on discrete wavelet transform (DWT) and singular value decomposition (SVD), leading to a 3D image hiding with high security and concealment. The encoding of POH and the utilization of CRPM can substantially enhance the level of security, and the DWT-SVD-based digital image watermarking can effectively hide the information of the 3D plaintext image in a visible ciphertext image, thus improving the imperceptibility of valid information. It is worth noting that the adopted structured light during the POH encoding possesses many optical parameters, which are all served as the supplementary keys, bringing about a great expansion of key space; meanwhile, the sensitivities of the wavelength key and singular matrix keys are also substantially enhanced thanks to the introduction of structured light, contributing to a significant enhancement of security. Numerical simulations are performed to demonstrate the feasibility of the proposed 3D image hiding method, and the simulation results show that the proposed method exhibits high feasibility and apparent security-enhanced effect as well as strong robustness. Full article

Multi-focus image fusion (MFIF) is an image-processing method that aims to generate fully focused images by integrating source images from different focal planes. However, the defocus spread effect (DSE) often leads to blurred or jagged focus/defocus boundaries in fused images, which affects the quality of the image. To address this issue, this paper proposes a novel model that embeds the Kolmogorov–Arnold network with convolutional layers in parallel within the U-Net architecture (KCUNet). This model keeps the spatial dimensions of the feature map constant to maintain high-resolution details while progressively increasing the number of channels to capture multi-level features at the encoding stage. In addition, KCUNet incorporates a content-guided attention mechanism to enhance edge information processing, which is crucial for DSE reduction and edge preservation. The model’s performance is optimized through a hybrid loss function that evaluates in several aspects, including edge alignment, mask prediction, and image quality. Finally, comparative evaluations against 15 state-of-the-art methods demonstrate KCUNet’s superior performance in both qualitative and quantitative analyses. Full article

This study presents a U-Net-based automatic segmentation framework for quantitative analysis of surface morphology in a PEEK-based composite following tribological testing. Controlled Pin-on-Disc tests were conducted to characterize tribological performance, worn surfaces were captured by laser scanning microscopy to acquire optical images and height maps, and the model produced pixel-level segmentation masks distinguishing different regions, enabling high-throughput, objective analysis of worn surface morphology. Sixty-three manually annotated image sets—with labels for fiber, third-body patch, and matrix regions—formed the training corpus. A 70-layer U-Net architecture with four-channel input was developed and rigorously evaluated using five-fold cross-validation. To enhance performance on the challenging patch and fiber classes, the top five model instances were ensembled through Bayesian-optimized weighted voting, achieving significant improvements in class-specific F1 metrics. Segmentation outputs on unseen data confirmed the method’s robustness and generalizability across complex surface topographies. This approach establishes a scalable, accurate tool for automated morphological analysis, with potential extensions to real-time monitoring and other composite systems. Full article

Crop disease identification is a pivotal research area in smart agriculture, forming the foundation for disease mapping and targeted prevention strategies. Among the most prevalent global wheat diseases, powdery mildew—caused by fungal infection—poses a significant threat to crop yield and quality, making early and accurate detection crucial for effective management. In this study, we present QY-SE-MResNet34, a deep learning-based classification model that builds upon ResNet34 to perform multi-class classification of wheat leaf images and assess powdery mildew severity at the single-leaf level. The proposed methodology begins with dataset construction following the GBT 17980.22-2000 national standard for powdery mildew severity grading, resulting in a curated collection of 4248 wheat leaf images at the grain-filling stage across six severity levels. To enhance model performance, we integrated transfer learning with ResNet34, leveraging pretrained weights to improve feature extraction and accelerate convergence. Further refinements included embedding a Squeeze-and-Excitation (SE) block to strengthen feature representation while maintaining computational efficiency. The model architecture was also optimized by modifying the first convolutional layer (conv1)—replacing the original 7 × 7 kernel with a 3 × 3 kernel, adjusting the stride to 1, and setting padding to 1—to better capture fine-grained leaf textures and edge features. Subsequently, the optimal training strategy was determined through hyperparameter tuning experiments, and GrabCut-based background processing along with data augmentation were introduced to enhance model robustness. In addition, interpretability techniques such as channel masking and Grad-CAM were employed to visualize the model’s decision-making process. Experimental validation demonstrated that QY-SE-MResNet34 achieved an 89% classification accuracy, outperforming established models such as ResNet50, VGG16, and MobileNetV2 and surpassing the original ResNet34 by 11%. This study delivers a high-performance solution for single-leaf wheat powdery mildew severity assessment, offering practical value for intelligent disease monitoring and early warning systems in precision agriculture. Full article

In recent years, the rapid advancement of deep learning techniques has significantly propelled the development of face forgery methods, drawing considerable attention to face forgery detection. However, existing detection methods still struggle with generalization across different datasets and forgery techniques. In this work, we address this challenge by leveraging both local texture cues and global frequency domain information in a complementary manner to enhance the robustness of face forgery detection. Specifically, we introduce a local texture mining and enhancement module. The input image is segmented into patches and a subset is strategically masked, then texture enhanced. This joint masking and enhancement strategy forces the model to focus on generalizable localized texture traces, mitigates overfitting to specific identity features and enabling the model to capture more meaningful subtle traces of forgery. Additionally, we extract multi-scale frequency domain features from the face image using wavelet transform, thereby preserving various frequency domain characteristics of the image. And we propose an innovative frequency-domain processing strategy to adjust the contributions of different frequency-domain components through frequency-domain selection and dynamic weighting. This Facilitates the model’s ability to uncover frequency-domain inconsistencies across various global frequency layers. Furthermore, we propose an integrated framework that combines these two feature modalities, enhanced with spatial attention and channel attention mechanisms, to foster a synergistic effect. Extensive experiments conducted on several benchmark datasets demonstrate that the proposed technique demonstrates superior performance and generalization capabilities compared to existing methods. Full article

Gait recognition is a reliable biometric approach that uniquely identifies individuals based on their natural walking patterns. It is widely used to recognize individuals who are challenging to camouflage and do not require a person’s cooperation. The general face-based person recognition system often fails to determine the offender’s identity when they conceal their face by wearing helmets and masks to evade identification. In such cases, gait-based recognition is ideal for identifying offenders, and most existing work leverages a deep learning (DL) model. However, a single model often fails to capture a comprehensive selection of refined patterns in input data when external factors are present, such as variation in viewing angle, clothing, and carrying conditions. In response to this, this paper introduces a fusion-based multi-model gait recognition framework that leverages the potential of convolutional neural networks (CNNs) and a vision transformer (ViT) in an ensemble manner to enhance gait recognition performance. Here, CNNs capture spatiotemporal features, and ViT features multiple attention layers that focus on a particular region of the gait image. The first step in this framework is to obtain the Gait Energy Image (GEI) by averaging a height-normalized gait silhouette sequence over a gait cycle, which can handle the left–right gait symmetry of the gait. After that, the GEI image is fed through multiple pre-trained models and fine-tuned precisely to extract the depth spatiotemporal feature. Later, three separate fusion strategies are conducted, and the first one is decision-level fusion (DLF), which takes each model’s decision and employs majority voting for the final decision. The second is feature-level fusion (FLF), which combines the features from individual models through pointwise addition before performing gait recognition. Finally, a hybrid fusion combines DLF and FLF for gait recognition. The performance of the multi-model fusion-based framework was evaluated on three publicly available gait databases: CASIA-B, OU-ISIR D, and the OU-ISIR Large Population dataset. The experimental results demonstrate that the fusion-enhanced framework achieves superior performance. Full article

The Anderson localization of phonons in disordered superlattices has been proposed as a route to suppress thermal conductivity beyond the limits imposed by conventional scattering mechanisms. A commonly used signature of phonon localization is the emergence of the nonmonotonic dependence of thermal conductivity $\kappa$ on system length L, i.e., a $\kappa$-L maximum. However, such behavior has rarely been observed. In this work, we conduct extensive non-equilibrium molecular dynamics (NEMD) simulations, using the LAMMPS package, on both periodic superlattices (SLs) and aperiodic random multilayers (RMLs) constructed from Si/Ge and Lennard-Jones materials. By systematically varying acoustic contrast, interatomic bond strength, and average layer thickness, we examine the interplay between coherent and incoherent phonon transport in these systems. Our two-phonon model decomposition reveals that coherent phonons alone consistently exhibit a strong nonmonotonic $\kappa$-L. This localization signature is often masked by the diffusive, monotonically increasing contribution from incoherent phonons. We further extract the ballistic-limit mean free paths for both phonon types, and demonstrate that incoherent transport often dominates, thereby concealing localization effects. Our findings highlight the importance of decoupling coherent and incoherent phonon contributions in both simulations and experiments. This work provides new insights and design principles for achieving phonon Anderson localization in superlattice structures. Full article

Modern industrial ecosystems, particularly those embracing Industry 4.0, increasingly depend on coherent optical networks operating at 400 Gbps and beyond. These high-capacity infrastructures, coupled with advanced digital signal processing and phase-sensitive detection, enable real-time data exchange for automated manufacturing, robotics, and interconnected factory systems. However, they introduce multilayer security challenges—ranging from hardware synchronization gaps to protocol overhead manipulation. Moreover, the rise of large-scale quantum computing intensifies these threats by potentially breaking classical key exchange protocols and enabling the future decryption of stored ciphertext. In this paper, we present a systematic vulnerability analysis of coherent optical networks that use OTU4 framing, Media Access Control Security (MACsec), and 400G ZR+ transceivers. Guided by established risk assessment methodologies, we uncover critical weaknesses affecting management plane interfaces (e.g., MDIO and I²C) and overhead fields (e.g., Trail Trace Identifier, Bit Interleaved Parity). To mitigate these risks while preserving the robust data throughput and low-latency demands of industrial automation, we propose a post-quantum security framework that merges spectral phase masking with multi-homodyne coherent detection, strengthened by quantum key distribution for key management. This layered approach maintains backward compatibility with existing infrastructure and ensures forward secrecy against quantum-enabled adversaries. The evaluation results show a substantial reduction in exposure to timing-based exploits, overhead field abuses, and cryptographic compromise. By integrating quantum-safe measures at the optical layer, our solution provides a future-proof roadmap for network operators, hardware vendors, and Industry 4.0 stakeholders tasked with safeguarding next-generation manufacturing and engineering processes. Full article

Crack detection in complex pavement scenarios remains challenging due to the sparse small-target features and computational inefficiency of existing methods. To address these limitations, this study proposes an enhanced architecture based on Mask2Former. The framework integrates two key innovations. A Local Perception Module (LPM) reconstructs geometric topological relationships through a Sequence-Space Dynamic Transformation Mechanism (DS2M), enhancing neighborhood feature extraction via depthwise separable convolutions. Simultaneously, an Auxiliary Convolutional Layer (ACL) combines lightweight residual convolutions with shallow high-resolution features, preserving critical edge details through channel attention weighting. Experimental evaluations demonstrate the model’s superior performance, achieving improvements of 3.2% in mIoU and 2.7% in mAcc compared to baseline methods, while maintaining computational efficiency with only 12.8 GFLOPs. These results validate the effectiveness of geometric relationship modeling and hierarchical feature fusion for pavement crack detection, suggesting practical potential for infrastructure maintenance systems. The proposed approach balances precision and efficiency, offering a viable solution for real-world applications with complex crack patterns and hardware constraints. Full article

We propose a secure and scalable file-encryption scheme for cloud systems by integrating Post-Quantum Cryptography (PQC), Quantum Key Distribution (QKD), and Advanced Encryption Standard (AES) within a distributed architecture. While prior studies have primarily focused on secure key exchange or authentication protocols (e.g., layered PQC-QKD key distribution), our scheme extends beyond key management by implementing a distributed encryption architecture that protects large-scale files through integrated PQC, QKD, and AES. To support high-throughput encryption, our proposed scheme partitions the target file into fixed-size subsets and distributes them across slave nodes, each performing parallel AES encryption using a locally reconstructed key from a PQC ciphertext. Each slave node receives a PQC ciphertext that encapsulates the AES key, along with a PQC secret key masked using QKD based on the BB84 protocol, both of which are centrally generated and managed by the master node for secure coordination. In addition, an encryption and transmission pipeline is designed to overlap I/O, encryption, and communication, thereby reducing idle time and improving resource utilization. The master node performs centralized decryption by collecting encrypted subsets, recovering the AES key, and executing decryption in parallel. Our evaluation using a real-world medical dataset shows that the proposed scheme achieves up to 2.37× speedup in end-to-end runtime and up to 8.11× speedup in encryption time over AES (Original). In addition to performance gains, our proposed scheme maintains low communication cost, stable CPU utilization across distributed nodes, and negligible overhead from quantum key management. Full article

Inverse lithography technology (ILT) based on the gradient descent (GD) algorithm, which is a classical local optimal method, can effectively improve the lithographic imaging fidelity. However, due to the low-pass filtering effect of the lithography imaging system, GD, although able to converge quickly, is prone to fall into the local optimum for the information in the corner region of complex patterns. Considering the high-frequency information of the corner region during the optimization process, this paper proposes a resolution layering method to improve the efficiency of GD-based ILT algorithms. A corner-rounding-inspired target retargeting strategy is used to compensate for the over-optimization defect of GD for inversely optimizing the complex pattern layout. Furthermore, for ensuring the manufacturability of masks, differentiable top-hat and bottom-hat operations are employed to improve the optimization efficiency of the proposed method. To confirm the superiority of the proposed method, multiple optimization methods of ILT were compared. Numerical experiments show that the proposed method has higher optimization efficiency and effectively avoids the over-optimization. Full article

With the advance of Artificial Intelligence, Deep Neural Networks are widely employed in various sensor-based systems to analyze operational conditions. However, due to the inherently nondeterministic and probabilistic natures of neural networks, the assurance of overall system performance could become a challenging task. In particular, soft errors could weaken the robustness of such networks and thereby threaten the system’s safety. Conventional fault-tolerant techniques by means of hardware redundancy and software correction mechanisms often involve a tricky trade-off between effectiveness and scalability in addressing the extensive design space of Deep Neural Networks. In this work, we propose a Reinforcement-Learning-based approach to protect neural networks from soft errors by addressing and identifying the vulnerable bits. The approach consists of three key steps: (1) analyzing layer-wise resiliency of Deep Neural Networks by a fault injection simulation; (2) generating layer-wise bit masks by a Reinforcement-Learning-based agent to reveal the vulnerable bits and to protect against them; and (3) synthesizing and deploying bit masks across the network with guaranteed operation efficiency by adopting transfer learning. As a case study, we select several existing neural networks to test and validate the design. The performance of the proposed approach is compared with the performance of other baseline methods, including Hamming code and the Most Significant Bits protection schemes. The results indicate that the proposed method exhibits a significant improvement. Specifically, we observe that the proposed method achieves a significant performance gain of at least 10% to 15% over on the test network. The results indicate that the proposed method dynamically and efficiently protects the vulnerable bits compared with the baseline methods. Full article

The accurate delineation of shrub crown boundaries is critical for ecological monitoring, land management, and understanding vegetation dynamics in fragile ecosystems such as semi-arid shrublands. While traditional image processing techniques often struggle with overlapping canopies, deep learning methods, such as convolutional neural networks (CNNs), offer promising solutions for precise segmentation. This study employed high-resolution imagery captured by unmanned aircraft systems (UASs) throughout the shrub growing season and explored the effectiveness of transfer learning for both semantic segmentation (Attention U-Net) and instance segmentation (Mask R-CNN). It utilized pre-trained model weights from two previous studies that originally focused on tree crown delineation to improve shrub crown segmentation in non-forested areas. Results showed that transfer learning alone did not achieve satisfactory performance due to differences in object characteristics and environmental conditions. However, fine-tuning the pre-trained models by unfreezing additional layers improved segmentation accuracy by around 30%. Fine-tuned pre-trained models show limited sensitivity to shrubs in the early growing season (April to June) and improved performance when shrub crowns become more spectrally unique in late summer (July to September). These findings highlight the value of combining pre-trained models with targeted fine-tuning to enhance model adaptability in complex remote sensing environments. The proposed framework demonstrates a scalable solution for ecological monitoring in data-scarce regions, supporting informed land management decisions and advancing the use of deep learning for long-term environmental monitoring. Full article

Infrared small target detection (IRSTD) aims to locate and separate targets from complex backgrounds. The challenges in IRSTD primarily come from extremely sparse target features and strong background clutter interference. However, existing methods typically perform discrimination directly on the features extracted by deep networks, neglecting the distinct characteristics of weak and small targets in the frequency domain, thereby limiting the improvement of detection capability. In this paper, we propose a frequency-aware masked-attention network (FM-Net) that leverages multi-scale frequency clues to assist in representing global context and suppressing noise interference. Specifically, we design the wavelet residual block (WRB) to extract multi-scale spatial and frequency features, which introduces a wavelet pyramid as the intermediate layer of the residual block. Then, to perceive global information on the long-range skip connections, a frequency-modulation masked-attention module (FMM) is used to interact with multi-layer features from the encoder. FMM contains two crucial elements: (a) a mask attention (MA) mechanism for injecting broad contextual feature efficiently to promote full-level semantic correlation and focus on salient regions, and (b) a channel-wise frequency modulation module (CFM) for enhancing the most informative frequency components and suppressing useless ones. Extensive experiments on three benchmark datasets (e.g., SIRST, NUDT-SIRST, IRSTD-1k) demonstrate that FM-Net achieves superior detection performance. Full article

Specific emitter identification (SEI) is a core technology for wireless device security that plays a crucial role in protecting wireless communication systems from various security threats. However, current deep learning-based SEI methods heavily rely on large amounts of labeled data for supervised training, facing challenges in non-cooperative communication scenarios. To address these issues, this paper proposes a novel contrastive asymmetric masked learning-based SEI (CAML-SEI) method, effectively solving the problem of SEI under scarce labeled samples. The proposed method constructs an asymmetric auto-encoder architecture, comprising an encoder network based on channel squeeze-and-excitation residual blocks to capture radio frequency fingerprint (RFF) features embedded in signals, while employing a lightweight single-layer convolutional decoder for masked signal reconstruction. This design promotes the learning of fine-grained local feature representations. To further enhance feature discriminability, a learnable non-linear mapping is introduced to compress high-dimensional encoded features into a compact low-dimensional space, accompanied by a contrastive loss function that simultaneously achieves feature aggregation of positive samples and feature separation of negative samples. Finally, the network is jointly optimized by combining signal reconstruction and feature contrast tasks. Experiments conducted on real-world ADS-B and Wi-Fi datasets demonstrate that the proposed method effectively learns generalized RFF features, and the results show superior performance compared with other SEI methods. Full article