Search Results (758)

Fine-grained food image segmentation is essential for accurate dietary assessment and nutritional analysis, yet remains highly challenging due to ambiguous boundaries, inter-class similarity, and dense layouts of meals containing many different ingredients in real-world settings. Existing methods based solely on CNNs, Transformers, or Mamba architectures often fail to simultaneously preserve fine-grained local details and capture contextual dependencies over long distances. To address these limitations, we propose HDF (Hybrid Decoder for Food Image Segmentation), a novel decoding framework built upon the MambaVision backbone. Our approach first employs a convolution-based feature pyramid network (FPN) to extract multi-stage features from the encoder. These features are then thoroughly fused across scales using a Cross-Layer Mamba module that models inter-level dependencies with linear complexity. Subsequently, an Attention Refinement module integrates global semantic context through spatial–channel reweighting. Finally, a Food Co-occurrence Module explicitly enhances food-specific semantics by learning dynamic co-occurrence patterns among categories, improving segmentation of visually similar or frequently co-occurring ingredients. Evaluated on two widely used, high-quality benchmarks, FoodSeg103 and UEC-FoodPIX Complete, which are standard datasets for fine-grained food segmentation, HDF achieves a 52.25% mean Intersection-over-Union (mIoU) on FoodSeg103 and a 76.16% mIoU on UEC-FoodPIX Complete, outperforming current state-of-the-art methods by a clear margin. These results demonstrate that HDF’s hybrid design and explicit co-occurrence awareness effectively address key challenges in food image segmentation, providing a robust foundation for practical applications in dietary logging, nutritional estimation, and food safety inspection. Full article

Small target detection in Unmanned Aerial Vehicle (UAV) applications is often plagued by inherent challenges such as small object sizes, sparse information, and complex background interference. Traditional detection algorithms and existing YOLO series models suffer from limitations in detection accuracy and fine-grained detail preservation. To address this, this paper proposes YOLO-DSNet, a small target detection network based on YOLOv13n. First, we introduce the dual-stream attention module (DSAM), which enhances discriminative features by leveraging bidirectional context modeling. Second, we design the Multi-scale Attention C2f (MSA-C2f) module—an adaptive architecture that optimizes feature extraction via multi-scale enhancement, effectively preserving and integrating small target information. Finally, through dataset augmentation, we significantly improve the model’s detection performance. The proposed YOLO-DSNet achieves a mAP@0.5 improvement from 30.8% to 40.1% on the VisDrone2019 dataset with only 0.8 million additional parameters, yielding a 30% accuracy gain while increasing computational overhead by merely 11.6 Gigaflops (GFLOPs). Experiments demonstrate YOLO-DSNet’s effectiveness in small target detection tasks such as UAV aerial photography and remote sensing imagery, successfully balancing accuracy and efficiency with high practical value. Full article

The rapid digitization of library resources has intensified the need for robust digital-rights management (DRM) mechanisms to safeguard copyright, control access, and preserve user privacy. Conventional DRM approaches are often centralized, prone to single-point-of-failure, and are limited in transparency and interoperability. To address these challenges, this article puts forward a decentralized DRM framework for library systems by leveraging blockchain technology and decentralized DRM-key mechanisms. An integrative review of the available research literature provides an analysis of current blockchain-based DRM library systems, their limitations, and associated challenges. To address these issues, a controlled experiment is set up to implement and evaluate a possible solution. In the proposed model, digital content is encrypted and stored in the Inter-Planetary File System (IPFS), while blockchain smart contracts manage the generation, distribution, and validation of DRM-keys that regulate user-access rights. This approach ensures immutability, transparency, and fine-grained access control without reliance on centralized authorities. Security is enhanced through cryptographic techniques for authentication. The model not only mitigates issues of piracy, unauthorized redistribution, and vendor lock-in, but also provides a scalable and interoperable solution for modern digital libraries. The findings demonstrate how blockchain-enabled DRM-keys can enhance trust, accountability, and efficiency through the development of secure, decentralized, and user-centric digital library systems, which will be of interest to practitioners charged with library IT technology management and to researchers in the wider field of blockchain applications in organizations. Full article

Object recognition in remote sensing images is essential for applications such as land resource monitoring, maritime vessel detection, and emergency disaster assessment. However, detection accuracy is often limited by complex backgrounds, densely distributed targets, and multiscale variations. To address these challenges, this study aims to improve the detection of small-scale and densely distributed objects in complex remote sensing scenes. An improved object detection network is proposed, called omnidirectional and adaptive YOLOv8 (OA-YOLOv8), based on the YOLOv8 architecture. Two targeted enhancements are introduced. First, an omnidirectional perception refinement (OPR) network is embedded into the backbone to strengthen multiscale feature representation through the incorporation of receptive-field convolution with a triplet attention mechanism. Second, an adaptive channel dynamic upsampling (ACDU) module is designed by combining DySample, the Haar wavelet transform, and a self-supervised equivariant attention mechanism (SEAM) to dynamically optimize channel information and preserve fine-grained features during upsampling. Experiments on the satellite imagery multi-vehicle dataset (SIMD) demonstrate that OA-YOLOv8 outperforms the original YOLOv8 by 4.6%, 6.7%, and 4.1% in terms of mAP@0.5, precision, and recall, respectively. Visualization results further confirm its superior performance in detecting small and dense targets, indicating strong potential for practical remote sensing applications. Full article

Wireless Sensor Networks (WSNs) are widely deployed for long-term monitoring in environments characterized by nonstationary sensing dynamics, intermittent connectivity and continuously evolving network topologies, while reliable, fine-grained labeled data capturing faults and adversarial behaviors remain scarce. This survey systematically reviews and synthesizes recent research that integrates autoencoder-based representation learning with self-supervised learning (SSL) objectives to enhance anomaly detection under these practical constraints. We structure the existing literature through a unified taxonomy encompassing autoencoder variants, self-supervised pretext tasks, spatio-temporal encoding mechanisms and the increasing use of graph-structured autoencoders for topology-aware modeling. Across distinct methodological categories, SSL-augmented frameworks consistently demonstrate improved robustness and stability compared to purely reconstruction-driven baselines, particularly in heterogeneous, dynamic and temporally drifting WSN environments. Nevertheless, this review also highlights several unresolved challenges that hinder real-world adoption, including uncertain scalability to large-scale networks, limited model interpretability, nontrivial energy and memory overheads on resource-constrained sensor nodes and a lack of standardized evaluation protocols and reporting practices. By consolidating publicly available datasets, experimental configurations and comparative performance trends, we derive concrete design requirements for robust and resource-aware anomaly detection in operational WSNs and outline promising future research directions, emphasizing lightweight model architectures, explainable learning mechanisms and federated AE–SSL paradigms to enable adaptive, privacy-preserving monitoring in next-generation IoT sensing systems. Full article

Fine-Grained Visual Classification (FGVC) involves distinguishing highly similar subordinate categories within the same basic-level class, presenting significant challenges due to subtle inter-class variations and substantial intra-class diversity. While Vision Transformer (ViT)-based approaches have demonstrated potential in this domain, they remain limited by two key issues: (1) the progressive loss of gradient-based edge and texture signals during hierarchical token aggregation and (2) insufficient extraction of discriminative fine-grained features. To overcome these limitations, we propose a Gradient-Aware Token Injection Transformer, a novel framework that explicitly incorporates gradient magnitude and orientation into token embeddings. This multi-modal feature fusion mechanism enhances the model’s capacity to preserve and leverage critical fine-grained visual cues. Extensive experiments on four standard FGVC benchmarks demonstrate the superiority of our approach, achieving 92.9% top-1 accuracy on CUB-200-2011, 90.5% on iNaturalist 2018, 93.2% on NABirds, and 95.3% on Stanford Cars, thereby validating its effectiveness and robustness. Full article

Water scarcity in arid/semi-arid regions restricts agricultural sustainability systems and hinders the achievement of regional sustainable development goals, especially in northwest China’s extremely arid areas, where acute water supply–demand conflicts and inefficient traditional practices intensify competition for water between agricultural and ecological sectors. This study aims to verify the effectiveness of an intelligent automatic irrigation system in mitigating water scarcity pressures and enhancing agricultural sustainability in the Shule River Basin of northwestern China, a region where traditional irrigation methods not only yield suboptimal crop outputs but also undermine long-term water resource sustainability. A smart irrigation module, integrating “sensing–decision–execution” processes, was embedded within a digital twin platform to enable precise, resource-efficient water management that aligns with sustainable development principles. Sunflower (Helianthus annuus L.), the most popular cash crop in the area, was used as the test crop, with three soil moisture-based irrigation levels compared against traditional farmer practices. Key indicators including leaf area index (LAI), dry biomass, grain yield, and irrigation water use efficiency (IWUE) were systematically evaluated. The results showed that (1) LAI increased from the seedling to flowering stage, with smart irrigation treatments significantly outperforming farmer practices in both crop growth and water-saving effects, laying a foundation for sustainable yield improvement; (2) total dry biomass at maturity was positively correlated with irrigation amount but smart irrigation optimized the allocation of water resources to avoid waste, balancing productivity and sustainability; (3) grain yield peaked within 70–89% field capacity (fc), with further increases leading to diminishing returns and unnecessary water consumption that impairs sustainable water use; (4) IWUE followed a parabolic trend, reaching its maximum under the same optimal irrigation range, indicating that smart irrigation can maximize water productivity while preserving water resources for ecological and future agricultural needs. The digital twin-driven smart irrigation system enhances both crop yield and water productivity in arid regions, providing a scalable model for precision water management in water-stressed agricultural zones. The results provide a key empirical basis and technical approach for sustainably using irrigation water, optimizing water–energy–food–ecology synergy, and advancing sustainable agriculture in arid regions of Northwest China, which is crucial for achieving regional sustainable development objectives amid worsening water scarcity. Full article

Accurate segmentation of 3D dental structures is essential for oral diagnosis, orthodontic planning, and digital dentistry. With the rapid advancement of 3D scanning and modeling technologies, high-resolution dental data have become increasingly common. However, existing approaches still struggle to process such high-resolution data efficiently. Current models often suffer from excessive parameter counts, slow inference, high computational overhead, and substantial GPU memory usage. These limitations compel many studies to downsample the input data to reduce training and inference costs—an operation that inevitably diminishes critical geometric details, blurs tooth boundaries, and compromises both fine-grained structural accuracy and model robustness. To address these challenges, this study proposes DiffusionNet++, an end-to-end segmentation framework capable of operating directly on raw high-resolution dental data. Building upon the standard DiffusionNet architecture, our method introduces a normal-enhanced multi-feature input strategy together with a lightweight SE channel-attention mechanism, enabling the model to effectively exploit local directional cues, curvature variations, and other higher-order geometric attributes while adaptively emphasizing discriminative feature channels. Experimental results demonstrate that the coordinates + normal feature configuration consistently delivers the best performance. DiffusionNet++ achieves substantial improvements in overall accuracy (OA), mean Intersection over Union (mIoU), and individual class IoU across all data types, while maintaining strong robustness and generalization on challenging cases, such as missing teeth and partially scanned data. Qualitative visualizations further corroborate these findings, showing superior boundary consistency, finer structural preservation, and enhanced recovery of incomplete regions. Overall, DiffusionNet++ offers an efficient, stable, and highly accurate solution for high-resolution 3D tooth segmentation, providing a powerful foundation for automated digital dentistry research and real-world clinical applications. Full article

The in-image machine translation task involves translating text embedded within images, with the translated results presented in image format. While this task has numerous applications in various scenarios such as film poster translation and everyday scene image translation, existing methods frequently neglect the aspect of consistency throughout this process. We propose the need to uphold two types of consistency in this task: translation consistency and image generation consistency. The former entails incorporating image information during translation, while the latter involves maintaining consistency between the style of the text image and the original image, ensuring background coherence. To address these consistency requirements, we introduce a novel two-stage framework named HCIIT (High-Consistency In-Image Translation), which involves text image translation using a multimodal multilingual large language model in the first stage and image backfilling with a diffusion model in the second stage. Chain-of-thought learning is employed in the first stage to enhance the model’s ability to effectively leverage visual information during translation. Subsequently, a diffusion model trained for style-consistent text–image generation is adopted. We further modify the structural network of the conventional diffusion model by introducing a style latent module, which ensures uniformity of text style within images while preserving fine-grained background details. The results obtained on both curated test sets and authentic image test sets validate the effectiveness of our framework in ensuring consistency and producing high-quality translated images. Full article

Face unmasking is a critical task in image restoration, as masks conceal essential facial features like the mouth, nose, and chin. Current inpainting methods often struggle with structural fidelity when handling large-area occlusions, leading to blurred or inconsistent results. To address this gap, we propose the Masked-to-Unmasked Network (M2UNet), a segmentation-guided generative framework. M2UNet leverages a segmentation-derived mask prior to accurately localize occluded regions and employs a multi-scale, attention-enhanced generator to restore fine-grained facial textures. The framework focuses on producing visually and semantically plausible reconstructions that preserve the structural logic of the face. Evaluated on a synthetic masked-face dataset derived from CelebA, M2UNet achieves state-of-the-art performance with a PSNR of 31.3375 dB and an SSIM of 0.9576. These results significantly outperform recent inpainting methods while maintaining high computational efficiency. Full article

Human activity recognition in privacy-sensitive indoor environments requires sensing modalities that remain robust under illumination variation and background clutter while preserving user anonymity. To this end, this study proposes a lightweight radar–camera fusion deep learning model that integrates motion signatures from FMCW radar with coarse spatial cues from ultra-low-resolution camera frames. The radar stream is processed as a Range–Doppler–Time cube, where each frame is flattened and sequentially encoded using a Transformer-based temporal model to capture fine-grained micro-Doppler patterns. The visual stream employs a privacy-preserving $4\times 5$-pixel camera input, from which a temporal sequence of difference frames is extracted and modeled with a dedicated camera Transformer encoder. The two modality-specific feature vectors—each representing the temporal dynamics of motion—are concatenated and passed through a lightweight fully connected classifier to predict human activity categories. A multimodal dataset of synchronized radar cubes and ultra-low-resolution camera sequences across 15 activity classes was constructed for evaluation. Experimental results show that the proposed fusion model achieves 98.74% classification accuracy, significantly outperforming single-modality baselines (single-radar and single-camera). Despite its performance, the entire model requires only 11 million floating-point operations (11 MFLOPs), making it highly efficient for deployment on embedded or edge devices. Full article

Precise identification of weeds at the maize seedling stage is pivotal for implementing Site-Specific Weed Management and minimizing herbicide environmental pollution. However, the performance of existing lightweight detectors is severely bottlenecked by unstructured field environments, characterized by the “green-on-green” spectral similarity between crops and weeds, diminutive seedling targets, and complex mutual occlusion of leaves. To address these challenges, this study proposes SLD-YOLO11, a topology-reconstructed lightweight detection model tailored for complex field environments. First, to mitigate the feature loss of tiny targets, a Lossless Downsampling Topology based on Space-to-Depth Convolution (SPD-Conv) is constructed, transforming spatial information into depth channels to preserve fine-grained features. Second, a Decomposed Large Kernel Attention (D-LKA) mechanism is designed to mimic the wide receptive field of human vision. By modeling long-range spatial dependencies with decomposed large-kernel attention, it enhances discrimination under severe occlusion by leveraging global structural context. Third, the DySample operator is introduced to replace static interpolation, enabling content-aware feature flow reconstruction. Experimental results demonstrate that SLD-YOLO11 achieves an mAP@0.5 of 97.4% on a self-collected maize field dataset, significantly outperforming YOLOv8n, YOLOv10n, YOLOv11n, and mainstream lightweight variants. Notably, the model achieves Zero Inter-class Misclassification between maize and weeds, establishing high safety standards for weeding operations. To further bridge the gap between visual perception and precision operations, a Visual Weed-Crop Competition Index (VWCI) is innovatively proposed. By integrating detection bounding boxes with species-specific morphological correction coefficients, the VWCI quantifies field weed pressure with low cost and high throughput. Regression analysis reveals a high consistency (R² = 0.70) between the automated VWCI and manual ground-truth coverage. This study not only provides a robust detector but also offers a reliable decision-making basis for real-time variable-rate spraying by intelligent weeding robots. Full article

To address the challenges of accurately extracting features from elongated scratches, irregular defects, and small-scale surface flaws on high-precision linear guide rails, this paper proposes a novel instance segmentation algorithm tailored for guide rail surface defect detection. The algorithm integrates the YOLOv8 instance segmentation framework with deformable convolutional networks and multi-scale feature fusion to enhance defect feature extraction and segmentation performance. A dedicated guide rail surface Defect (GSD) segmentation dataset is constructed to support model training and evaluation. In the backbone, the DCNv3 module is incorporated to strengthen the extraction of elongated and irregular defect features while simultaneously reducing model parameters. In the feature fusion network, a multi-scale feature fusion module and a triple-feature encoding module are introduced to jointly capture global contextual information and preserve fine-grained local defect details. Furthermore, a Channel and Position Attention Module (CPAM) is employed to integrate global and local features, improving the model’s sensitivity to channel and positional cues of small-target defects and thereby enhancing segmentation accuracy. Experimental results show that, compared with the original YOLOv8n-Seg, the proposed method achieves improvements of 3.9% and 3.8% in Box and Mask mAP₅₀, while maintaining a real-time inference speed of 148 FPS. Additional evaluations on the public MSD dataset further demonstrate the model’s strong versatility and robustness. Full article

Coastal lakes are highly vulnerable transitional systems in which sedimentological processes and benthic ecological conditions jointly control contaminant accumulation and preservation, particularly in densely urbanized settings. A robust understanding of the physical and ecological characteristics of bottom sediments is therefore essential for the correct interpretation of contaminant distributions, including those of potentially toxic metals. In this study, an integrated sedimentological–ecological approach was applied to Lake Ganzirri, a Mediterranean shallow coastal lake located in northeastern Sicily (Italy), where recent investigations have identified localized heavy metal anomalies in surface sediments. Sediment texture, petrographic and mineralogical composition, malacofaunal assemblages, and lake-floor morpho-bathymetry were systematically analysed using grain-size statistics, faunistic determinations, GIS-based spatial mapping, and bivariate and multivariate statistical methods. The modern lake bottom is dominated by bioclastic quartzo-lithic sands with low fine-grained fractions and variable but locally high contents of calcareous skeletal remains, mainly derived from molluscs. Sediments are texturally heterogeneous, consisting predominantly of coarse-grained sands with lenses of very coarse sand, along with gravel and subordinate medium-grained sands. Both sedimentological features and malacofaunal death assemblages indicate deposition under open-lagoon conditions characterized by brackish waters and relatively high hydrodynamic energy. Spatial comparison between sedimentological–ecological parameters and previously published heavy metal distributions reveals no significant correlations with metal hotspots. The generally low metal concentrations, mostly below regulatory threshold values, are interpreted as being favoured by the high permeability and mobility of coarse sediments and by energetic hydrodynamic conditions limiting fine-particle accumulation. Overall, the integration of sedimentological and ecological data provides a robust framework for interpreting contaminant patterns and offers valuable insights for the environmental assessment and management of vulnerable coastal lake systems, as well as for the understanding of modern lagoonal sedimentary processes. Full article

To address the challenges of compressing large-scale, multi-channel temperature data from aluminum alloy heating furnaces—and the limitations of traditional methods in preserving fidelity for critical tasks like energy accounting and process playback—this paper proposes an elastic, task-aware time-series compression method based on semantic segmentation. The method automatically segments data and annotates anchor points according to key process stages and significant operational events. Data are grouped by furnace number and alloy grade into segment-level buckets. Within this structure, an enhanced PCA model is built using channel-specific weights and amplified anchor points. The optimal principal component dimension is selected automatically under explained variance constraints, with channel-wise DCT used as a fallback for small samples. Compression accuracy is evaluated using combined rRMSE metrics (overall and per temperature channel) and key event recall rate. Experiments show the method achieves an average overall rRMSE of 0.11624, a temperature channel rRMSE of 0.08860, and a compression ratio of 1.18, outperforming Standard-PCA, PAA, and RP-Gauss. Notably, the proposed method achieves 100% recall for key events during heat preservation, demonstrating superior performance. Further analysis shows performance varies significantly across process stages, furnace IDs, and alloy grades, offering valuable insights for fine-grained evaluation and real-world deployment. Full article