Explore the vast range of titles available.

Fault transient impulses induced by faulty components in rotating machinery usually contain substantial interference. Fault features are comparatively weak in the initial fault stage, which renders fault diagnosis more difficult. In this case, a sparse representation method based on the Majorzation-Minimization (MM) algorithm is proposed to enhance weak fault features and extract the features from strong background noise. However, the traditional MM algorithm suffers from two issues, which are the choice of sparse basis and complicated calculations. To address these challenges, a modified MM algorithm is proposed in which a sparse optimization objective function is designed firstly. Inspired by the Basis Pursuit (BP) model, the optimization function integrates an impulsive feature-preserving factor and a penalty function factor. Second, a modified Majorization iterative method is applied to address the convex optimization problem of the designed function. A series of sparse coefficients can be achieved through iterating, which only contain transient components. It is noteworthy that there is no need to select the sparse basis in the proposed iterative method because it is fixed as a unit matrix. Then the reconstruction step is omitted, which can significantly increase detection efficiency. Eventually, envelope analysis of the sparse coefficients is performed to extract weak fault features. Simulated and experimental signals including bearings and gearboxes are employed to validate the effectiveness of the proposed method. In addition, comparisons are made to prove that the proposed method outperforms the traditional MM algorithm in terms of detection results and efficiency.

The complex and harsh working environment of rolling bearings cause the fault characteristics in vibration signal contaminated by the noise, which make fault diagnosis difficult. In this paper, a feature enhancement method of rolling bearing signal based on variational mode decomposition with K determined adaptively (K-adaptive VMD), and radial based function fuzzy entropy (RBF-FuzzyEn), is proposed. Firstly, a phenomenon called abnormal decline of center frequency (ADCF) is defined in order to determine the parameter K of VMD adaptively. Then, the raw signal is separated into K intrinsic mode functions (IMFs). A coefficient En for selecting optimal IMFs is calculated based on the center frequency bands (CFBs) of all IMFs and frequency spectrum for original signal autocorrelation operation. After that, the optimal IMFs of which En are bigger than the threshold are selected to reconstruct signal. Secondly, RBF is introduced as an innovative fuzzy function to enhance the feature discrimination of fuzzy entropy between bearings in different states. A specific way for determination of parameter r in fuzzy function is also presented. Finally, RBF-FuzzyEn is used to extract features of reconstructed signal. Simulation and experiment results show that K-adaptive VMD can effectively reduce the noise and enhance the fault characteristics; RBF-FuzzyEn has strong feature differentiation, superior noise robustness, and low dependence on data length.

Sensor-based activity recognition (AR) depends on effective feature representation and classification. However, many recent studies focus on recognition methods, but largely ignore feature representation. Benefitting from the success of Convolutional Neural Networks (CNN) in feature extraction, we propose to improve the feature representation of activities. Specifically, we use a reversed CNN to generate the significant data based on the original features and combine the raw training data with significant data to obtain to enhanced training data. The proposed method can not only train better feature extractors but also help better understand the abstract features of sensor-based activity data. To demonstrate the effectiveness of our proposed method, we conduct comparative experiments with CNN Classifier and CNN-LSTM Classifier on five public datasets, namely the UCIHAR, UniMiB SHAR, OPPORTUNITY, WISDM, and PAMAP2. In addition, we evaluate our proposed method in comparison with traditional methods such as Decision Tree, Multi-layer Perceptron, Extremely randomized trees, Random Forest, and k-Nearest Neighbour on a specific dataset, WISDM. The results show our proposed method consistently outperforms the state-of-the-art methods.

Optical coherence tomography (OCT) acquisitions often reduce lateral sampling density to shorten scan time and suppress motion artifacts, but this strategy degrades the signal-to-noise ratio and obscures fine retinal microstructures. To recover these details without hardware modifications, we propose MGFormer, a lightweight Transformer for OCT super-resolution (SR) that integrates a multi-granularity attention mechanism with tensor distillation. A feature-enhancing convolution first sharpens edges; stacked multi-granularity attention blocks then fuse coarse-to-fine context, while a row-wise top-k operator retains the most informative tokens and preserves their positional order. We trained and evaluated MGFormer on B-scans from the Duke SD-OCT dataset at 2×, 4×, and 8× scaling factors. Relative to seven recent CNN- and Transformer-based SR models, MGFormer achieves the highest quantitative fidelity; at 4× it reaches 34.39 dB PSNR and 0.8399 SSIM, surpassing SwinIR by +0.52 dB and +0.026 SSIM, and reduces LPIPS by 21.4%. Compared with the same backbone without tensor distillation, FLOPs drop from 289G to 233G (−19.4%), and per-B-scan latency at 4× falls from 166.43 ms to 98.17 ms (−41.01%); the model size remains compact (105.68 MB). A blinded reader study shows higher scores for boundary sharpness (4.2 ± 0.3), pathology discernibility (4.1 ± 0.3), and diagnostic confidence (4.3 ± 0.2), exceeding SwinIR by 0.3–0.5 points. These results suggest that MGFormer can provide fast, high-fidelity OCT SR suitable for routine clinical workflows.

Plant species, structural combination, and spatial distribution in different regions should be adapted to local conditions, and the reasonable arrangement can bring the best ecological effect. Therefore, it is essential to understand the classification and distribution of plant species. This paper proposed an end-to-end network with Enhancing Nested Downsampling features (END-Net) to solve complex and challenging plant species segmentation tasks. There are two meaningful operations in the proposed network: (1) A compact and complete encoder–decoder structure nests in the down-sampling process; it makes each downsampling block obtain the equal feature size of input and output to get more in-depth plant species information. (2) The downsampling process of the encoder–decoder framework adopts a novel pixel-based enhance module. The enhanced module adaptively enhances each pixel’s features with the designed learnable variable map, which is as large as the corresponding feature map and has n×n variables; it can capture and enhance each pixel’s information flexibly effectively. In the experiments, our END-Net compared with eleven state-of-the-art semantic segmentation architectures on the self-collected dataset, it has the best PA (Pixel Accuracy) score and FWloU (Frequency Weighted Intersection over Union) accuracy and achieves 84.52% and 74.96%, respectively. END-Net is a lightweight model with excellent performance; it is practical in complex vegetation distribution with aerial and optical images. END-Net has the following merits: (1) The proposed enhancing module utilizes the learnable variable map to enhance features of each pixel adaptively. (2) We nest a tiny encoder–decoder module into the downsampling block to obtain the in-depth plant species features with the same scale in- and out-features. (3) We embed the enhancing module into the nested model to enhance and extract distinct plant species features. (4) We construct a specific plant dataset that collects the optical images-based plant picture captured by drone with sixteen species.

Image dehazing is an important problem since computer recognition requires high-quality inputs. Recently, many researches tend to build an end-to-end multiscale network to restore haze-free images. But unfortunately, existing multiscale networks tend to recover under-dehazed results due to inefficient feature extraction. To solve the problem, we propose an enhanced context aggregation network for single image dehazing named ECANet. Based on encoder–decoder structure, the ECANet improves feature representation by three feature aggregation blocks (FABs) on each scale. The FAB is a new efficient feature extraction module, which adequately extracts content features and style features due to the difference receptive field between dilated convolution and ordinary convolution. To better fuse these complementary features, we combine spatial and channel attention mechanism to each FAB. After the decoding process, we also adopt an enhancing block to further refine image details under the supervision of clear references. The experimental results show that the proposed ECANet performs better than state-of-the-art dehazing methods, which recovers clear images with discriminative texture and natural color.

In industrial production, defect detection for automotive headlight lenses is an essential yet challenging task. Transparent glass defect detection faces several difficulties, including a wide variety of defect shapes and sizes, as well as the challenge of identifying transparent surface defects. To enhance the accuracy and efficiency of this process, we propose a computer vision-based inspection solution utilizing multi-angle lighting. For this task, we collected 2000 automotive headlight images to systematically categorize defects in transparent glass, with the primary defect types being spots, scratches, and abrasions. During data acquisition, we proposed a dataset augmentation method named SWAM to address class imbalance, ultimately generating the Lens Defect Dataset (LDD), which comprises 5532 images across these three main defect categories. Furthermore, we propose a defect detection network named the Transparent Glass Defect Network (TGDNet), designed based on common transparent glass defect types. Within the backbone of TGDNet, we introduced the TGFE module to adaptively extract local features for different defect categories and employed TGD, an improved SK attention mechanism, combined with a spatial attention mechanism to boost the network's capability in multi-scale feature fusion. Experiments demonstrate that compared to other classical defect detection methods, TGDNet achieves superior performance on the LDD, improving the average detection precision by 6.7% in mAP and 8.9% in mAP50 over the highest-performing baseline algorithm.

Mini-batch partitioning is a widely used technique in deep learning that involves dividing a dataset into smaller subsets. This method is crucial in training deep learning models such as deep neural networks and convolutional neural networks. It is favored for its ability to accelerate model convergence, reduce memory overhead, and minimize convergence errors. The primary advantage of mini-batch partitioning is that it allows the model to learn dataset features more evenly, thereby speeding up the convergence process. However, determining the optimal method and size for mini-batch partitioning remains a challenging problem. This paper proposes a novel mini-batch partitioning method focused on feature reorganization. By leveraging a Self-Organizing Map network for feature extraction, data with similar characteristics are initially grouped into the same batch. The purity index of each batch is then calculated based on the number of features and labels, allowing for a comprehensive evaluation of batch homogeneity. Batches with significant differences in purity are selectively reorganized to ensure that each batch contains a diverse set of features, reducing intra-batch feature correlation and ultimately enhancing data representation.Furthermore, through SOM network mapping, the dataset can be effectively partitioned into subsets that are well-suited for model training. Experimental comparisons of various batch partitioning methods on multiple UCI datasets demonstrate that our proposed method, termed MPBS (Mini-Batch Partitioning Algorithm based on Self-Organizing Map Network). Compared with other algorithms, the accuracy, loss and training time are improved by 14.06%, 24.31% and 31.22%.

This paper focuses on the high-resolution (HR) remote sensing images semantic segmentation task, whose goal is to predict semantic labels in a pixel-wise manner. Due to the rich complexity and heterogeneity of information in HR remote sensing images, the ability to extract spatial details (boundary information) and semantic context information dominates the performance in segmentation. In this paper, based on the frequently used fully convolutional network framework, we propose a boundary enhancing semantic context network (BES-Net) to explicitly use the boundary to enhance semantic context extraction. BES-Net mainly consists of three modules: (1) a boundary extraction module for extracting the semantic boundary information, (2) a multi-scale semantic context fusion module for fusing semantic features containing objects with multiple scales, and (3) a boundary enhancing semantic context module for explicitly enhancing the fused semantic features with the extracted boundary information to improve the intra-class semantic consistency, especially in those pixels containing boundaries. Extensive experimental evaluations and comprehensive ablation studies on the ISPRS Vaihingen and Potsdam datasets demonstrate the effectiveness of BES-Net, yielding an overall improvement of 1.28/2.36/0.72 percent in mF1/mIoU/OA over FCN_8s when the BE and MSF modules are combined by the BES module. In particular, our BES-Net achieves a state-of-the-art performance of 91.4% OA on the ISPRS Vaihingen dataset and 92.9%/91.5% mF1/OA on the ISPRS Potsdam dataset.

Purpose To identify qualitative MRI features of non-(contrast)-enhancing tumor (nCET) in glioblastoma’s T2-FLAIR hyperintense lesion. Methods Thirty-three histologically confirmed glioblastoma patients whose T1-, T2- and contrast-enhanced T1-weighted MRI and 11 C-methionine positron emission tomography (Met-PET) were available were included in this study. Met-PET was utilized as a surrogate for tumor burden. Imaging features for identifying nCET were searched by qualitative examination of 156 targets. A new scoring system to identify nCET was established and validated by two independent observers. Results Three imaging features were found helpful for identifying nCET; “Bulky gray matter involvement”, “Around the rim of contrast-enhancement (Around-rim),” and “High-intensity on T1WI and low-intensity on T2WI (HighT1LowT2)” resulting in an nCET score = 2 × Bulky gray matter involvement – 2 × Around-rim + HighT1LowT2 + 2. The nCET score’s classification performances of two independent observers measured by AUC were 0.78 and 0.80, with sensitivities and specificities using a threshold of four being 0.443 and 0.771, and 0.916 and 0.768, respectively. The weighted kappa coefficient for the nCET score was 0.946. Conclusion The current investigation demonstrated that qualitative assessments of glioblastoma’s MRI might help identify nCET in T2/FLAIR high-intensity lesions. The novel nCET score is expected to aid in expanding treatment targets within the T2/FLAIR high-intensity lesions. Graphical abstract