Explore the vast range of titles available.

Accurate segmentation in medical imaging requires loss functions that capture both regional overlap and boundary alignment. This study evaluates composite losses combining binary cross-entropy (BCE) and a boundary-based term under fixed and adaptive weighting schemes, using U-Net and SwinUNETR on the MM-WHS dataset. For U-Net, a small boundary contribution with adaptive weighting yielded the best results: Standard SoftAdapt (90/10 BCE + BoundaryDoU) achieved the highest Dice score ( ), surpassing both the baseline ( ) and fixed ratios. In contrast, SwinUNETR achieved its strongest performance with a fixed 70% BCE + 10% boundary ratio (0.919 ± 0.02). The result showed that combining a boundary-based loss term helps improve the segmentation accuracy. However, the performance gain is dependent on the architecture of the segmentation model; convolution-based U-Net benefited from the adaptive loss weighting scheme, whereas Transformer-based SwinUNETR without strong inductive bias did not benefit from increased influence of the boundary loss term.

Railwayfreight cars operating in heavy-load and complex outdoor environments are frequently subject to adverse conditions such as haze, temperature fluctuations, and transmission interference, which significantly degrade the quality of the acquired images and introduce substantial noise. Furthermore, the structural complexity of freight cars, coupled with the small size, diversity, and complex structure of defect areas, poses serious challenges for image denoising. Specifically, it becomes extremely difficult to remove noise while simultaneously preserving fine-grained textures and edge details. These challenges distinguish railway freight car image denoising from conventional image restoration tasks, necessitating the design of specialized algorithms that can achieve both effective noise suppression and precise structural detail preservation. To address the challenges of incomplete denoising and poor preservation of details and edge information in railway freight car images, this paper proposes a novel image denoising algorithm named the Nonlinear Activation-Free Network based on Multi-Scale Edge Enhancement and Fusion (NAF-MEEF). The algorithm constructs a Multi-scale Edge Enhancement Initialization Layer to strengthen edge information at multiple scales. Additionally, it employs a Nonlinear Activation-Free feature extractor that effectively captures local and global image information. Leveraging the network’s multi-branch parallelism, a Multi-scale Rotation Fusion Attention Mechanism is developed to perform weight analysis on information across various scales and dimensions. To ensure consistency in image details and structure, this paper introduces a fusion loss function. The experimental results show that compared with recent advanced methods, the proposed algorithm has better noise suppression and edge preservation performance. The proposed method achieves significant denoising performance on railway freight car images affected by Gaussian, composite, and simulated real-world noise, with PSNR gains of 1.20 dB, 1.45 dB, and 0.69 dB, and SSIM improvements of 2.23%, 2.72%, and 1.08%, respectively. On public benchmarks, it attains average PSNRs of 30.34 dB (Set12) and 28.94 dB (BSD68), outperforming several state-of-the-art methods. In addition, this method also performs well in railway image dehazing tasks and demonstrates good generalization ability in denoising tests of remote sensing ship images, further proving its robustness and practical application value in diverse image restoration tasks.

What are the main findings? We propose a novel Multi-Attention Fusion Generative Adversarial Network (MAF-GAN) that integrates oriented convolution, multi-dimensional attention mechanisms, and dynamic feature fusion to significantly enhance the reconstruction of directional structures and fine textures in remote sensing imagery. Extensive experiments demonstrate that MAF-GAN achieves state-of-the-art performance on the GF7-SR4×-MSD dataset, with a PSNR of 27.14 dB and SSIM of 0.7206, outperforming existing mainstream models while maintaining a favorable balance between reconstruction quality and inference efficiency. What are the implication of the main findings? The proposed model provides a reliable and efficient technical pathway for generating high-resolution remote sensing images with clearer spatial structures and more natural spectral characteristics, supporting high-precision applications such as urban planning and environmental monitoring. The introduced modular design, including oriented convolution, multi-attention fusion, and a composite loss function, offers a flexible and extensible framework that can inspire future research in specialized super-resolution tasks for remote sensing and other geospatial image processing domains. Existing Generative Adversarial Networks (GANs) frequently yield remote sensing images with blurred fine details, distorted textures, and compromised spatial structures when applied to super-resolution (SR) tasks, so this study proposes a Multi-Attention Fusion Generative Adversarial Network (MAF-GAN) to address these limitations: the generator of MAF-GAN is built on a U-Net backbone, which incorporates Oriented Convolutions (OrientedConv) to enhance the extraction of directional features and textures, while a novel co-calibration mechanism—incorporating channel, spatial, gating, and spectral attention—is embedded in the encoding path and skip connections, supplemented by an adaptive weighting strategy to enable effective multi-scale feature fusion, and a composite loss function is further designed to integrate adversarial loss, perceptual loss, hybrid pixel loss, total variation loss, and feature consistency loss for optimizing model performance; extensive experiments on the GF7-SR4×-MSD dataset demonstrate that MAF-GAN achieves state-of-the-art performance, delivering a Peak Signal-to-Noise Ratio (PSNR) of 27.14 dB, Structural Similarity Index (SSIM) of 0.7206, Learned Perceptual Image Patch Similarity (LPIPS) of 0.1017, and Spectral Angle Mapper (SAM) of 1.0871, which significantly outperforms mainstream models including SRGAN, ESRGAN, SwinIR, HAT, and ESatSR as well as exceeds traditional interpolation methods (e.g., Bicubic) by a substantial margin, and notably, MAF-GAN maintains an excellent balance between reconstruction quality and inference efficiency to further reinforce its advantages over competing methods; additionally, ablation studies validate the individual contribution of each proposed component to the model’s overall performance, and this method generates super-resolution remote sensing images with more natural visual perception, clearer spatial structures, and superior spectral fidelity, thus offering a reliable technical solution for high-precision remote sensing applications.

This study proposes a new component of the composite loss function minimised during training of the Super-Resolution (SR) algorithms—the normalised structural similarity index loss , which has the potential to improve the natural appearance of reconstructed images. Deep learning-based super-resolution (SR) algorithms reconstruct high-resolution images from low-resolution inputs, offering a practical means to enhance image quality without requiring superior imaging hardware, which is particularly important in medical applications where diagnostic accuracy is critical. Although recent SR methods employing convolutional and generative adversarial networks achieve high pixel fidelity, visual artefacts may persist, making the design of the loss function during training essential for ensuring reliable and naturalistic image reconstruction. Our research shows on two models—SR and Invertible Rescaling Neural Network (IRN)—trained on multiple benchmark datasets that the function significantly contributes to the visual quality, preserving the structural fidelity on the reference datasets. The quantitative analysis of results while incorporating shows that including this loss function component has a mean 2.88% impact on the improvement of the final structural similarity of the reconstructed images in the validation set, in comparison to leaving it out and 0.218% in comparison when this component is non-normalised.

Infrared images of sensitive targets are difficult to obtain and cannot meet the design and training needs of target detection and tracking algorithms for mobile platforms such as aircraft. This paper proposes an image translation algorithm TransImg, which can achieve visible light image translation to the infrared domain to enrich the dataset. First, the algorithm designed a generator structure consisting of a deep residual connected encoder and a region perception feature fusion module to enhance feature learning, thereby avoiding issues such as generating infrared images with insufficient details in the transfer task. Afterward, a multi-scale discriminator and a composite loss function were designed to further improve the transfer effect. Finally, an automatic mixed-precision training strategy was designed for the overall migration algorithm architecture to accelerate the training and generation of infrared images. Experiments have shown that the image translation algorithm TransImg has good algorithm accuracy, and the infrared image generated by visible light image translation has richer texture details, faster generation speed, and lower video memory consumption, and the performance exceeds the mainstream traditional algorithm, and the generated images can meet the requirements of target detection and tracking algorithms design and training for mobile platforms such as aircraft.

The spatial QRS-T angle is a promising health indicator for risk stratification of sudden cardiac death (SCD). Thus far, the angle is estimated solely from 12-lead electrocardiogram (ECG) systems uncomfortable for ambulatory monitoring. Methods to estimate QRS-T angles from reduced-lead ECGs registered with consumer healthcare devices would, therefore, facilitate ambulatory monitoring. (1) Objective: Develop a method to estimate spatial QRS-T angles from reduced-lead ECGs. (2) Approach: We designed a deep learning model to locate the QRS and T wave vectors necessary for computing the QRS-T angle. We implemented an original loss function to guide the model in the 3D space to search for each vector’s coordinates. A gradual reduction of ECG leads from the largest publicly available dataset of clinical 12-lead ECG recordings (PTB-XL) is used for training and validation. (3) Results: The spatial QRS-T angle can be estimated from leads I, II, aVF, V2 with sufficient accuracy (absolute mean and median errors of 11.4° and 7.3°) for detecting abnormal angles without sacrificing patient comfortability. (4) Significance: Our model could enable ambulatory monitoring of spatial QRS-T angles using patch- or textile-based ECG devices. Populations at risk of SCD, like chronic cardiac and kidney disease patients, might benefit from this technology.

Three-dimensional gravity inversion technology involves inferring the underground density structure based on observed gravity anomaly data. In addition to gravity inversion based on physics-driven methods, deep learning, as a purely data-driven technique, is increasingly gaining attention in geophysical inversion problems. However, purely data-driven methods rely on the implicit relationships within the data during the inversion process, which results in a lack of clear physical significance. This study proposes a three-dimensional gravity inversion method that integrates physical equations with deep learning. Based on the U-Net architecture, the gravity forward equation is incorporated as a physical constraint term, and a composite loss function—comprising three-dimensional mean squared error, a depth-weighting function, and three-dimensional intersection-over-union loss—is constructed to enhance inversion accuracy. Numerical experiments indicate that this method outperforms traditional algorithms in terms of density recovery accuracy and boundary clarity. When applied to gravity anomaly data from the Tangshan earthquake region in China, this method successfully inverted the three-dimensional subsurface density structure, revealing a high-density anomaly beneath the seismic source area, which provides important evidence for understanding the regional earthquake generation mechanism.

Distributed Denial-of-Service (DDoS) attacks generate deceptive, high-volume traffic that bypasses conventional detection mechanisms. When interception fails, effectively allocating mixed benign and malicious traffic under resource constraints becomes a critical challenge. To address this, we propose SchedOpt Engine (SOE), a scheduling framework formulated as a discrete multi-objective optimization problem. The goal is to optimize four conflicting objectives: a benign traffic acceptance rate (BTAR), malicious traffic interception rate (MTIR), server load balancing, and malicious traffic isolation. These objectives are combined into a composite scalarized loss function with soft constraints, prioritizing a BTAR while maintaining flexibility. To solve this problem, we introduce MOFATA, a multi-objective extension of the Fata Morgana Algorithm (FATA) within a Pareto-based evolutionary framework. An ϵ-dominance mechanism is incorporated to improve solution granularity and diversity. Simulations under varying attack intensities and resource constraints validate the effectiveness of SOE. Results show that SOE consistently achieves a high BTAR and MTIR while balancing server loads. Under extreme attacks, SOE isolates malicious traffic to a subset of servers, preserving capacity for benign services. SOE also demonstrates strong adaptability in fluctuating attack environments, providing a practical solution for DDoS mitigation.

PurposeMulti-domain convolutional neural network (MDCNN) model has been widely used in object recognition and tracking in the field of computer vision. However, if the objects to be tracked move rapid or the appearances of moving objects vary dramatically, the conventional MDCNN model will suffer from the model drift problem. To solve such problem in tracking rapid objects under limiting environment for MDCNN model, this paper proposed an auto-attentional mechanism-based MDCNN (AA-MDCNN) model for the rapid moving and changing objects tracking under limiting environment.Design/methodology/approachFirst, to distinguish the foreground object between background and other similar objects, the auto-attentional mechanism is used to selectively aggregate the weighted summation of all feature maps to make the similar features related to each other. Then, the bidirectional gated recurrent unit (Bi-GRU) architecture is used to integrate all the feature maps to selectively emphasize the importance of the correlated feature maps. Finally, the final feature map is obtained by fusion the above two feature maps for object tracking. In addition, a composite loss function is constructed to solve the similar but different attribute sequences tracking using conventional MDCNN model.FindingsIn order to validate the effectiveness and feasibility of the proposed AA-MDCNN model, this paper used ImageNet-Vid dataset to train the object tracking model, and the OTB-50 dataset is used to validate the AA-MDCNN tracking model. Experimental results have shown that the augmentation of auto-attentional mechanism will improve the accuracy rate 2.75% and success rate 2.41%, respectively. In addition, the authors also selected six complex tracking scenarios in OTB-50 dataset; over eleven attributes have been validated that the proposed AA-MDCNN model outperformed than the comparative models over nine attributes. In addition, except for the scenario of multi-objects moving with each other, the proposed AA-MDCNN model solved the majority rapid moving objects tracking scenarios and outperformed than the comparative models on such complex scenarios.Originality/valueThis paper introduced the auto-attentional mechanism into MDCNN model and adopted Bi-GRU architecture to extract key features. By using the proposed AA-MDCNN model, rapid object tracking under complex background, motion blur and occlusion objects has better effect, and such model is expected to be further applied to the rapid object tracking in the real world.

In many problems, one has several models of interest that capture key parameters describing the distribution of the data. Partially overlapping models are taken as models in which at least one covariate effect is common to the models. A priori knowledge of such structure enables efficient estimation of all model parameters. However, in practice, this structure may be unknown. We propose adaptive composite M-estimation (ACME) for partially overlapping models using a composite loss function, which is a linear combination of loss functions defining the individual models. Penalization is applied to pairwise differences of parameters across models, resulting in data driven identification of the overlap structure. Further penalization is imposed on the individual parameters, enabling sparse estimation in the regression setting. The recovery of the overlap structure enables more efficient parameter estimation. An oracle result is established. Simulation studies illustrate the advantages of ACME over existing methods that fit individual models separately or make strong a priori assumption about the overlap structure.