Sensors最新文献

Soil organic matter (SOM) is a key indicator for evaluating soil fertility and ecological functions, and hyperspectral technology provides an effective means for its rapid and non-destructive estimation. However, in practical soil systems, the spectral response of SOM is often highly covariant with mineral composition, moisture conditions, and soil structural characteristics. Under small-sample conditions, hyperspectral SOM modeling results are usually highly sensitive to spectral preprocessing methods, sample perturbations, and model architecture and parameter configurations, leading to fluctuations in predictive performance across independent runs and thereby limiting model stability and practical applicability. To address these issues, this study proposes a multi-strategy collaborative deep learning modeling framework for small-sample conditions (SE-EDCNN-DA-LWGPSO). Under unified data partitioning and evaluation settings, the framework integrates spectral preprocessing, data augmentation based on sensor perturbation simulation, multi-scale dilated convolution feature extraction, an SE channel attention mechanism, and a linearly weighted generalized particle swarm optimization algorithm. Subtropical red soil samples from Guangxi were used as the study object. Samples were partitioned using the SPXY method, and multiple independent repeated experiments were conducted to evaluate the predictive performance and training consistency of the model under fixed validation conditions. The results indicate that the combination of Savitzky-Golay filtering and first-derivative transformation (SG-1DR) exhibits superior overall stability among various preprocessing schemes. In model structure comparison and ablation analysis, as dilated convolution, data augmentation, and channel attention mechanisms were progressively introduced, the fluctuations of prediction errors on the validation set gradually converged, and the performance dispersion among different independent runs was significantly reduced. Under ten independent repeated experiments, the final model achieved R² = 0.938 ± 0.010, RMSE = 2.256 ± 0.176 g·kg^-1, and RPD = 4.050 ± 0.305 on the validation set, demonstrating that the proposed framework has good modeling consistency and numerical stability under small-sample conditions.

This paper presents the optimization of surface modification for aptameric graphene nanosensors for the measurement of biomarkers in undiluted physiological media. In these sensors, graphene transduces the binding between an aptamer and the intended target biomarker into a measurable signal while being coated with a polyethylene glycol (PEG) nanolayer to minimize nonspecific adsorption of matrix molecules. We perform a systematic study of the aptamer and PEG attachment schemes and parameters, including the impact of the serial or parallel PEG-aptamer attachment scheme, PEG molecular weight and surface density, and aptamer surface density on the sensor behavior, such as the responsivity to biomarker concentration changes, and importantly, they are used for operation in physiological media and have the ability to reject nonspecific binding to interfering molecules. We then use the understanding from this parametric study to identify graphene nanosensor designs that are optimally functionalized with PEG and aptamers to be strongly responsive to target biomarkers and effectively reduce nonspecific adsorption of interferents, thereby enabling sensitive and specific biomarker measurements in undiluted physiological media. The experimental results show that nanosensors that were optimized via serial modification with 5000 Da PEG at 15 mM and a 94 nt DNA aptamer at 500 nM allowed specific measurement of C-reactive protein (CRP) in undiluted human serum with a limit of detection (LOD) down to 27 pM, representing an up to 1000-fold improvement compared to previously reported CRP measurements.

While deep learning methods are increasingly applied in the field of DOA estimation, existing approaches generally feed the real and imaginary parts of the covariance matrix directly into neural networks without optimizing the input features, which prevents classical attention mechanisms from improving accuracy. This paper proposes a spatio-temporal fusion model named DACL-Net for DOA estimation. The spatial branch applies a two-dimensional Fourier transform (2D-FT) to the covariance matrix, causing angles to appear as peaks in the magnitude spectrum. This operation transforms the original covariance matrix into a dark image with bright spots, enabling the convolutional neural network (CNN) to focus on the bright-spot components via an attention module. Additionally, a spectrum attention mechanism (SAM) is introduced to enhance the extraction of temporal features in the time branch. The model learns simultaneously from two data branches and finally outputs DOA results through a linear layer. Simulation results demonstrate that DACL-Net outperforms existing algorithms in terms of accuracy, achieving an RMSE of 0.04° at an SNR of 0 dB.

Understanding human movement in industrial environments requires more than simple step counts-it demands contextual information to interpret activities and enhance workflows. Key factors such as location and process context are essential. However, research on context-sensitive human activity recognition is limited by the lack of publicly available datasets that include both human movement and contextual labels. Our work introduces the DaRA dataset to address this research gap. DaRA comprises over 109 h of video footage, including 32 h from wearable first-person cameras and 77 h from fixed third-person cameras. In a laboratory environment replicating a realistic warehouse, scenarios such as order picking, packaging, unpacking, and storage were captured. The movements of 18 subjects were captured using inertial measurement units, Bluetooth devices for indoor localization, wearable first-person cameras, and fixed third-person cameras. DaRA offers detailed annotations with 12 class categories and 207 class labels covering human movements and contextual information such as process steps and locations. A total of 15 annotators and 8 revisers contributed over 1572 h in annotation and 361 h in revision. High label quality is reflected in Light's Kappa values ranging from 78.27% to 99.88%. Therefore, DaRA provides a robust, multimodal foundation for human activity and context recognition in industrial settings.

Background: Upper limb loss has a profound impact on individuals' daily activities, self-image, and social interactions. Despite continuous technological advances in upper-limb prosthetics, high rates of device abandonment persist, highlighting the need to better understand users' functional and psychosocial needs.

Methods: To gain a deeper understanding of the perspectives of upper limb amputees and the synthesis of their needs across ergonomic, functional, and psychological dimensions, this study was conducted. A systematic review was conducted following PRISMA guidelines to synthesize user-reported evidence on upper-limb prosthesis use. Articles indexed in the Web of Science database between 2016 and December 2023 were screened using predefined search terms related to upper-limb amputation, prostheses, social impact, and user needs. Studies were included if they reported direct perspectives of upper-limb prosthesis users regarding usability, functionality, and lived experience.

Results: Out of 239 papers identified, 31 were included and analyzed. The findings reveal that functional performance, comfort, weight, intuitive control, and reliability are strongly interconnected with psychosocial factors such as confidence, embodiment, social participation, and acceptance. Technological advances have not consistently translated into improved alignment between prosthetic solutions and user needs, which is reflected in continued dissatisfaction and abandonment.

Conclusions: This review provides a structured synthesis of user-reported needs across functional, ergonomic, and psychosocial dimensions, translating these insights into design-relevant guidelines. Emphasizing a user-centered and interdisciplinary perspective, the findings aim to support the development of upper-limb prosthetic devices that are more usable, acceptable, and aligned with users' expectations, ultimately bridging the gap between user expectations and technological capabilities and promoting long-term adoption and quality of life.

Accurate cardiac output (CO) measurement is vital for hemodynamic management; however, it usually requires invasive monitoring, which limits its continuous and out-of-hospital use. Wearable sensors integrated with deep learning offer a noninvasive alternative. This study developed and validated a lightweight deep learning model using wearable electrocardiography (ECG) and photoplethysmography (PPG) signals to predict CO and examined whether cardiac index-based normalization (Cardiac Index (CI) = CO/body surface area) improves performance. Twenty-seven patients who underwent cardiac surgery and had pulmonary artery catheters were prospectively enrolled. Single-lead ECG (HiCardi+ chest patch) and finger PPG (WristOx2 3150) were recorded simultaneously and processed through an ECG-PPG fusion network with cross-modal interaction. Three models were trained as follows: (1) CI prediction, (2) direct CO prediction, and (3) indirect CO prediction. The total number of CO = predicted CI × body surface area. Reference values were derived from thermodilution. The CI model achieved the best performance, and the indirect CO model showed significant reductions in error/agreement metrics (MAE/RMSE/bias; p < 0.0001), while correlation-based metrics are reported descriptively without implying statistical significance. The Pearson correlation coefficient (PCC) and percentage error (PE) for the indirect CO estimates (PCC = 0.904; PE = 23.75%). The indirect CO estimates met the predefined PE < 30% agreement benchmark for method-comparison; this is not a universal clinical standard. These results demonstrate that wearable ECG-PPG fusion deep learning can achieve accurate, noninvasive CO estimation and that CI-based normalization enhances model agreement with pulmonary artery catheter measurements, supporting continuous catheter-free hemodynamic monitoring.

To address the challenges of insufficient efficiency and accuracy in traditional detection models caused by minute target sizes, low signal-to-noise ratios (SNRs), and feature volatility in Satellite Laser Ranging (SLR) images, this paper proposes an efficient, lightweight, and high-precision detection model. The core motivation of this study is to fundamentally enhance the model's capabilities in feature extraction, fusion, and localization for minute and blurred targets through a specifically designed network architecture and loss function, without significantly increasing the computational burden. To achieve this goal, we first design a DMS-Conv module. By employing dense sampling and channel function separation strategies, this module effectively expands the receptive field while avoiding the high computational overhead and sampling artifacts associated with traditional multi-scale methods, thereby significantly improving feature representation for faint targets. Secondly, to optimize information flow within the feature pyramid, we propose a Lightweight Upsampling Module (LUM). Integrating depthwise separable convolutions with a channel reshuffling mechanism, this module replaces traditional transposed convolutions at a minimal computational cost, facilitating more efficient multi-scale feature fusion. Finally, addressing the stringent requirements for small target localization accuracy, we introduce the MPD-IoU Loss. By incorporating the diagonal distance of bounding boxes as a geometric penalty term, this loss function provides finer and more direct spatial alignment constraints for model training, effectively boosting localization precision. Experimental results on a self-constructed real-world SLR observation dataset demonstrate that the proposed model achieves an mAP50:95 of 47.13% and an F1-score of 88.24%, with only 2.57 M parameters and 6.7 GFLOPs. Outperforming various mainstream lightweight detectors in the comprehensive performance of precision and recall, these results validate that our method effectively resolves the small target detection challenges in SLR scenarios while maintaining a lightweight design, exhibiting superior performance and practical value.

Semantic segmentation of high-resolution remote sensing imagery is a critical technology for the intelligent interpretation of sensor data, supporting automated environmental monitoring and urban sensing systems. However, processing data from dense urban scenarios remains challenging due to sensor signal occlusions (e.g., shadows) and the complexity of parsing multi-scale targets from optical sensors. Existing approaches often exhibit a trade-off between the accuracy of global semantic modeling and the precision of complex boundary recognition. While the Segment Anything Model (SAM) offers powerful zero-shot structural priors, its direct application to remote sensing is hindered by domain gaps and the lack of inherent semantic categorization. To address these limitations, we propose a dual-branch cooperative network, PriorSAM-DBNet. The main branch employs a Densely Connected Swin (DC-Swin) Transformer to capture cross-scale global features via a hierarchical shifted window attention mechanism. The auxiliary branch leverages SAM's zero-shot capability to exploit structural universality, generating object-boundary masks as robust signal priors while bypassing semantic domain shifts. Crucially, we introduce a parameter-efficient Scaled Subsampling Projection (SSP) module that employs a weight-sharing mechanism to align cross-modal features, freezing the massive SAM backbone to ensure computational viability for practical sensor applications. Furthermore, a novel Attentive Cross-Modal Fusion (ACMF) module is designed to dynamically resolve semantic ambiguities by calibrating the global context with local structural priors. Extensive experiments on the ISPRS Vaihingen, Potsdam, and LoveDA-Urban datasets demonstrate that PriorSAM-DBNet outperforms state-of-the-art approaches. By fine-tuning only 0.91 million parameters in the auxiliary branch, our method achieves mIoU scores of 82.50%, 85.59%, and 53.36%, respectively. The proposed framework offers a scalable, high-precision solution for remote sensing semantic segmentation, particularly effective for disaster emergency response where rapid feature recognition from sensor streams is paramount.

In order to solve the problem of false alarms and missed alarms in pipeline monitoring caused by a large amount of noise in the negative pressure wave signal collected by pressure sensors, a new pressure signal denoising method based on the black-winged kite algorithm (BWK) is proposed. First, the variational mode decomposition (VMD) parameters are optimized through BWK. Next, the effective modal components are screened by sample entropy, and the secondary noise reduction of the signal is carried out by using the wavelet thresholding (WT). Finally, the signal is reconstructed to achieve noise reduction. Simulation experiments show that, compared with WT and empirical mode decomposition (EMD), the method proposed in this paper can achieve the best noise reduction effect under both high and low signal-to-noise ratio (SNR) conditions. The method proposed in the paper can achieve the highest SNR of 14.2280 dB, compared to WT's SNR of 12.6458 dB and EMD's SNR of 5.5292 dB. To further validate the performance of the algorithm, an experimental platform for simulating pipeline leaks is built. Compared with WT and EMD, the method proposed in this paper also shows the best noise reduction effect. This method provides a high-precision and adaptive solution for leak detection in urban water supply pipelines and has strong engineering application value.

Cybersecurity in radio frequency (RF) technologies has become a critical concern, driven by the expansion of connected systems in urban and industrial environments. Although research on wireless networks and the Internet of Things (IoT) has advanced, comprehensive studies that provide a global and integrated view of cybersecurity development in this field remain limited. This work presents a scientometric and systematic review of international publications from 2009 to 2025, integrating the PRISMA protocol with semantic screening supported by a Large Language Model to enhance classification accuracy and reproducibility. The analysis identified two interdependent axes: one focusing on signal integrity and authentication in GNSS systems and cellular networks; the other addressing the resilience of IoT networks, both strongly associated with spoofing and jamming, as well as replay, relay, eavesdropping, and man-in-the-middle (MitM) attacks. The results highlight the relevance of RF cybersecurity in securing communication infrastructures and expose gaps in widely adopted technologies such as RFID, NFC, BLE, ZigBee, LoRa, Wi-Fi, and unlicensed ISM bands, as well as in emerging areas like terahertz and 6G. These gaps directly affect the reliability and availability of IoT and wireless communication systems, increasing security risks in large-scale deployments such as smart cities and cyber-physical infrastructures.