Extract Features Research Articles

Unmanned aerial vehicle aerial image stitching method based on superpixel segmentation

Addressing the issues of long processing time, high computational complexity, and poor stitching quality in existing methods for unmanned aerial vehicle (UAV) aerial image stitching, this paper proposes an aerial image stitching method based on similar region estimation. Firstly, the input set of UAV aerial images is subjected to superpixel segmentation to estimate similar regions among the images. Secondly, an improved SIFT algorithm is used to extract and match feature points within the similar regions across the image set. Finally, based on the obtained matching points, an improved optimal seam-line algorithm and a weighted average fusion algorithm are employed to fuse the images, resulting in the stitched aerial image. Experimental validation using the public dataset UAV-image-mosaicing dataset demonstrates that our method achieves nearly double the speed of feature extraction compared to traditional feature extraction algorithms and triples the feature matching rate. Furthermore, compared to mainstream UAV aerial image stitching methods, our approach reduces time consumption by half while improving the stitching image evaluation metrics of SSIM, MAE, and PSNR by approximately 5% on average.

Enhancing multiple sclerosis diagnosis: A comparative study of electroencephalogram signal processing and entropy methods.

As one of the most common neurodegenerative diseases, Multiple sclerosis (MS) is a chronic immune-driven disorder that affects the central nervous system (CNS). Due to the variety of symptoms, accurately diagnosing MS demands rigorous attention to differential diagnosis, as various disorders can closely mimic its clinical and paraclinical features. Although MR imaging techniques are gold standards in diagnosing MS, the feasibility of advanced Electroencephalogram (EEG) signal processing methods is discussed in this study to detect patients with MS disorder. EEG signals from 50 individuals were evaluated through entropy-based methods. Sixteen distinct entropy methods were employed to extract features, which were used to train several machine-learning algorithms for classifying MS patients. Furthermore, each entropy method was individually evaluated to identify the most effective approach for MS diagnosis. A regional analysis of the EEG channels was conducted to determine the most informative regions for classification. The results indicated that the proposed method outperformed previous studies and achieved highly effective results in the classification of MS patients.

ECCDN-Net: A deep learning-based technique for efficient organic and recyclable waste classification.

Efficient waste management is essential to minimizing environmental harm as well as encouraging sustainable progress. The escalating volume and sophistication of waste present significant challenges, prompting innovative methods for effective waste categorization and management. Deep learning models have become highly intriguing tools for automating trash categorization activities, providing effective ways to optimize processes for handling waste. Ourwork presents a novel deep learning method for trash classification, with the goal to improve the accuracy, also efficiency of garbage image categorization. We examined the effectiveness of several pre-trained models, such as InceptionV2, Densenet201, MobileNet v2, and Resnet18, using objective evaluation and cross-validation. We proposed an Eco Cycle Classifier Deep Neural Network (ECCDN-Net) model that is particularly built for the categorization of waste images. ECCDN-Net utilizes the advantageous qualities of Densenet201 and Resnet18 by merging their capacities to extract features, enhanced with auxiliary outputs to optimize the classification procedure. The set of imagesused in this study comprises 24,705 images that are divided into two distinct classes: Organic and Recyclable. The set allows extensive evaluation and training of deep learning models for waste classification of images tasks. Our research demonstrates that the ECCDN-Net model classifies waste images with 96.10% accuracy, outperforming other pre-trained models. Resnet18 had 92.68% accuracy, MobileNet v2 93.27%, Inception v3 94.77%, and Densenet201, a significant improvement, 95.98%. ECCDN-Net outperformed these models in waste image categorization with 96.10% accuracy. We ensure the reliability and generalizability of our methods throughout the dataset by integrating and cross-validating deep learning models. The current work introduces an innovative deep learning-based approach that has promising potential for waste categorization and management strategies.

Development of deep learning-based mobile application for the identification of Coccidia species in pigs using microscopic images.

Coccidiosis is a gastrointestinal parasitic disease caused by different species of Eimeria and Isospora, poses a significant threat to pig farming, leading to substantial economic losses attributed to reduced growth rates, poor feed conversion, increased mortality rates, and the expense of treatment. Traditional methods for identifying Coccidia species in pigs rely on fecal examination and microscopic analysis, necessitating expert personnel for accurate species identification. To address this need, a deep learning-based mobile application capable of automatically identifying different species of Eimeria and Isospora was developed. The present study focused on six species, namely, E. debliecki, E. perminuta, E. porci, E. spinosa, E. suis, and Isospora suis, commonly found in pigs of the North Eastern Hill (NEH) region of India. Utilizing a two-stage approach, segmentation of coccidia oocysts in microscopic images using convolutional neural networks (CNNs), followed by species identification by same network was carried out in this work. Resource-efficient models, including EfficientNetB0, EfficientNetB1, MobileNet, and MobileNetV2, within an encoder-decoder architecture were utilized to extract features. Transfer learning was applied to enhance model accuracy during training. Additionally, a marker-controlled watershed algorithm was implemented to separate touching cells, thus reducing misclassification. The results demonstrate that all the developed models effectively segmented/classified Coccidia species, achieving mean Intersection-over-Union (m-IoU) values exceeding 0.92, with individual class IoU scores above 0.90. MobileNetV2 exhibited the highest m-IoU of 0.95, followed by EfficientNetB1 with an m-IoU of 0.94. For classification, MobileNetV2 demonstrated the highest performance, with accuracy, precision, and recall values of 0.93, 0.96, and 0.96, respectively. EfficientNetB1 yielded an accuracy of 0.91. The developed mobile application, tested on new data, achieved an identification accuracy of 91.0 %. These findings highlight the potential of deep learning-based mobile applications in effectively identifying Coccidia species in pigs, thus, providing a promising solution to mitigate reliance on expert personnel and laborious time-consuming experiments in this domain.

Pre-screening of endomysial microvessel density by fast random forest image processing machine learning algorithm accelerates recognition of a modified vascular network in idiopathic inflammatory myopathies

Biomarkers for discrimination among different subgroups of idiopathic inflammatory myopathies (IIM) are difficult to identify and may involve multiple laboratory tests and time-consuming procedures. We assessed the potential for artificial intelligence (AI) to extract features such as density of endomysial microvessels based on automatic analysis of the CD31+ vascular network on muscle biopsy images. We also assessed the potential of this technique to save time and its agreement rate with analyses based on the manual selection of microvessels from the same images. A total of 84 images from 84 patients with IIM, diagnosed between 2014 and 2020, were retrieved and analyzed using the Fast Random Forest (FRF) technique. We built a lightweight and explainable algorithm for calculating the pixel percentage of CD31+ endomysial capillaries. The FRF technique applied on images of CD31-stained muscle sections achieved a good performance in the recognition of microvessels by estimating their density over a standard area corresponding to a sample of microscope image. The time spent for this analysis was 90% less than the manual choice of microvessels (estimated time considering the computational time and the time spent to manually detecting the microvessels features). The good performance of the FRF demonstrates that the CD31 pixel percentage of endomysial capillaries is sufficient for a correct estimation. Finally, the paper proposes a procedure to integrate AI in the pre-screening process.

Improved lung nodule segmentation with a squeeze excitation dilated attention based residual UNet.

The diverse types and sizes, proximity to non-nodule structures, identical shape characteristics, and varying sizes of nodules make them challenging for segmentation methods. Although many efforts have been made in automatic lung nodule segmentation, most of them have not sufficiently addressed the challenges related to the type and size of nodules, such as juxta-pleural and juxta-vascular nodules. The current research introduces a Squeeze-Excitation Dilated Attention-based Residual U-Net (SEDARU-Net) with a robust intensity normalization technique to address the challenges related to different types and sizes of lung nodules and to achieve an improved lung nodule segmentation. After preprocessing the images with the intensity normalization method and extracting the Regions of Interest by YOLOv3, they are fed into the SEDARU-Net with dilated convolutions in the encoder part. Then, the extracted features are given to the decoder part, which involves transposed convolutions, Squeeze-Excitation Dilated Residual blocks, and skip connections equipped with an Attention Gate, to decode the feature maps and construct the segmentation mask. The proposed model was evaluated using the publicly available Lung Nodule Analysis 2016 (LUNA16) dataset, achieving a Dice Similarity Coefficient of 97.86%, IoU of 96.40%, sensitivity of 96.54%, and precision of 98.84%. Finally, it was shown that each added component to the U-Net's structure and the intensity normalization technique increased the Dice Similarity Coefficient by more than 2%. The proposed method suggests a potential clinical tool to address challenges related to the segmentation of lung nodules with different types located in the proximity of non-nodule structures.

Semi-supervised learning-based identification of the attachment between sludge and microparticles in wastewater treatment.

Monitoring the microparticle transfer process in wastewater treatment systems is crucial for improving treatment performance. Supervised deep learning methods show high performance to automatically detect particles, but they rely on vast amounts of labeled data for training. To overcome this issue, we proposed a semi-supervised learning (SSL) method based on the Simple framework for Contrastive Learning of visual Representations (SimCLR), to detect microparticles free from sludge and attached to sludge. First, we pre-trained a ResNet50 backbone by SimCLR, to extract features from much unlabeled data (1,000 images). Then, we constructed a Mask R-CNN architecture based on the pre-trained ResNet50, and fine-tuned it on a small quantity of labeled data (≈200 images with ≈600 annotated particles) in supervised learning fashion. We showcased its performance and practical applicability for microscopy images obtained from the water lab of TU Delft. The results demonstrate that the SSL methods obtain a significant improvement in mean average precision of up to 5% compared to the conventional supervised learning method, when a limited amount of labeled data is available (e.g., 91 labeled images). Furthermore, these methods improve the average precision for detecting attached particles by over 12%. With the detection results from the SSL methods, we measured the attachment efficiency of microparticles to sludge under varying mixed liquor suspended solids concentration and aeration intensity. The precise measurements demonstrate the effectiveness and practical applicability of the SSL method in facilitating long-term monitoring of particle transfer processes in biological wastewater treatment systems.

A Comprehensive Strategy for the Identification of Arachnoid Cysts in the Brain Utilizing Image Processing Segmentation Methods

This study focuses on the segmentation and characterization of arachnoid cysts in brain MRI images, aiming to enhance diagnostic accuracy through advanced image processing techniques. Arachnoid cysts are cerebrospinal fluid-filled sacs located between the brain or spinal cord and the arachnoid membrane. These cysts can be asymptomatic but may also cause neurological symptoms such as headaches, seizures, or cognitive impairments when they increase in size or pressure. Accurate detection and characterization are essential for timely intervention and treatment. In this research, 269 brain MRI images were analyzed using connected component analysis (CCA) and contrast-limited adaptive histogram equalization (CLAHE). CLAHE was employed to enhance image contrast, particularly in regions with subtle intensity differences, while CCA facilitated the segmentation of connected regions corresponding to cysts. The smallest connected components were identified and analyzed to isolate arachnoid cysts with high precision. Post-segmentation, quantitative analysis was performed to extract features such as size, shape, and density, enabling comprehensive cyst characterization. Additionally, calculations for area and approximate volume were conducted, providing critical information for clinical assessment. Visual validation of segmentation outcomes confirmed the effectiveness of the applied methods in accurately delineating cyst boundaries. This research addresses a significant gap in the existing literature. While most studies focus on brain tumor segmentation, there is limited work on arachnoid cyst detection and volume estimation. By integrating image processing techniques tailored for arachnoid cysts, this study offers a novel approach to their diagnosis and monitoring. The findings demonstrate the potential for automated diagnostic tools, reducing subjectivity and improving efficiency in clinical workflows. The proposed methodology aligns with advancements in medical imaging and contributes to the development of improved tools for neuroimaging diagnostics, paving the way for more precise and reliable assessments in the detection of brain pathologies.

Early Leaf Disease Detection of Soybean Plants using Convolution Neural Network Algorithm

Background: Agricultural specialists usually examine leaves closely to check for plant diseases, that process takes a long time and can have errors. To overcome these problems, nowadays, machine learning methods, specifically sequential convolution neural networks (CNN) are widely used, because of their extensive potential to extract features and patterns from the image dataset. Methods: This study introduces a complete methodology for detecting diseases in soybean plants by employing Convolutional Neural Networks (CNNs). A dataset sourced from the Mendeley database, containing images of soybean legume crops affected by caterpillars, Diabrotica Speciosa diseases and undamaged (healthy) foliage is used., The research explores the efficacy of the CNN Model in accurately classifying these diseases. The CNN algorithm is trained in such a way that it can handle the complex nature of soybean leaf imaging by utilizing convolutional layers for extracting features, pooling layers for reducing dimensionality and softmax layers for classification. Result: The training and validation results exhibit remarkable accuracy rates (97.64% accuracy) after 150 epochs. The evaluation metrics, such as precision, recall and F1-score, demonstrate the model’s performance across different leaf diseases of soybean, suggesting its ability to accurately identify instances inside each category. The classification matrices provide complete knowledge of the accuracy of the prediction of different diseases. The overall accuracy of the proposed model is 94.05%. This study can be utilized as a reference to increase the progress of agricultural disease detection, in turn, can enable enhance the reduction of crop losses and ensure food security.

MutualDTA: An Interpretable Drug-Target Affinity Prediction Model Leveraging Pretrained Models and Mutual Attention.

Efficient and accurate drug-target affinity (DTA) prediction can significantly accelerate the drug development process. Recently, deep learning models have been widely applied to DTA prediction and have achieved notable success. However, existing methods often encounter several common issues: first, the data representations lack sufficient information; second, the extracted features are not comprehensive; and third, most methods lack interpretability when modeling drug-target binding. To overcome the above-mentioned problems, we propose an interpretable deep learning model called MutualDTA for predicting DTA. MutualDTA leverages the power of pretrained models to obtain accurate representations of drugs and targets. It also employs well-designed modules to extract hidden features from these representations. Furthermore, the interpretability of MutualDTA is realized by the Mutual-Attention module, which (i) establishes relationships between drugs and proteins from the perspective of intermolecular interactions between drug atoms and protein amino acid residues and (ii) allows MutualDTA to capture the binding sites based on attention scores. The test results on two benchmark data sets show that MutualDTA achieves the best performance compared to the 12 state-of-the-art models. Attention visualization experiments show that MutualDTA can capture partial interaction sites, which not only helps drug developers reduce the search space for binding sites, but also demonstrates the interpretability of MutualDTA. Finally, the trained MutualDTA is applied to screen high-affinity drug screens targeting Alzheimer's disease (AD)-related proteins, and the screened drugs are partially present in the anti-AD drug library. These results demonstrate the reliability of MutualDTA in drug development.

Automated Age Estimation from OPG Images and Patient Records Using Deep Feature Extraction and Modified Genetic–Random Forest

Background/Objectives: Dental age estimation is a vital component of forensic science, helping to determine the identity and actual age of an individual. However, its effectiveness is challenged by methodological variability and biological differences between individuals. Therefore, to overcome the drawbacks such as the dependence on manual measurements, requiring a lot of time and effort, and the difficulty of routine clinical application due to large sample sizes, we aimed to automatically estimate tooth age from panoramic radiographs (OPGs) using artificial intelligence (AI) algorithms. Methods: Two-Dimensional Deep Convolutional Neural Network (2D-DCNN) and One-Dimensional Deep Convolutional Neural Network (1D-DCNN) techniques were used to extract features from panoramic radiographs and patient records. To perform age estimation using feature information, Genetic algorithm (GA) and Random Forest algorithm (RF) were modified, combined, and defined as Modified Genetic–Random Forest Algorithm (MG-RF). The performance of the system used in our study was analyzed based on the MSE, MAE, RMSE, and R2 values calculated during the implementation of the code. Results: As a result of the applied algorithms, the MSE value was 0.00027, MAE value was 0.0079, RMSE was 0.0888, and R2 score was 0.999. Conclusions: The findings of our study indicate that the AI-based system employed herein is an effective tool for age detection. Consequently, we propose that this technology could be utilized in forensic sciences in the future.

A new cross-domain approach for bearing fault diagnosis based on multiscale convolutional networks and adversarial subdomain adaptation

ABSTRACT Fault diagnosis of rolling bearings is of paramount significance in the field of engineering, as it directly impacts the reliability and safety of mechanical systems. Although deep learning techniques have demonstrated promising performance in this field, their efficacy often diminishes under varying operational conditions or when labelled data is limited. To overcome these challenges, this paper introduces a more precise cross-domain approach for bearing fault diagnosis. The proposed method exploits the temporal multiscale characteristics of vibration signals and the inherent multiscale nature of bearing faults by constructing a one-dimensional multiscale convolutional neural network. This network extracts features at multiple scales from raw vibration signals, which are then fused to form robust and generalised representations. Additionally, the integration of the Efficient Channel Attention mechanism further refines feature selection, enhancing the overall performance of the model. The label classifier, in conjunction with the Nuclear-norm Wasserstein Discrepancy, serves as a domain discriminator to facilitate adversarial domain adaptation. Concurrently, local maximum mean discrepancy and adversarial domain adaptation techniques align both global and subdomain feature distributions. Furthermore, label smoothing is incorporated to enhance the model’s generalisation capabilities. Experimental validation on the CWRU and PU rotating machinery datasets demonstrates the method’s exceptional robustness and superior transferability.

Spectral Pyramid Pooling and Fused Keypoint Generation in ResNet-50 for Robust 3D Object Detection

Accurate and robust 3D object detection under various environments is a challenging task and 3D object detection mainly relies on the direction, position, and size of the objects. The traditional object detection approaches are affected based on diverse issues such as background clutter, camera view-point variations, and occlusions. To avoid these issues in the field of object detection, one of the novel approaches named as Faster Region-based Convolutional Neural ResNet-50 (FRCNResNet-50) model is proposed that detects and classifies the 3D objects in the images. At first, the images are collected from three different data sources such as KITTI dataset, a nuScene dataset, and an MIT Indoor Scene dataset. For enhancing generalization ability and image quality, the collected images are included into preprocessing stages. The ResNet-50 model is designed to extract features by using a Spectral Pyramid Pooling (SPP) layer and a Fused Keypoint Generation (FKG) layer that enhances detection efficiency and reduces computational cost. The FRCNN model is implemented to detect 3D objects that include the ROI pooling layer for multi-class classification and for presenting its corresponding regression bounding box. The experimental validation is performed based on the significant measurements and quantitative analyses that showed the proposed model achieved better performances of 98.58% from accuracy analysis and 98 ms from computational time.

Research on Financial Market Prediction Model Based on Big Data and Quantum Computing

A financial market is a trading system for financial assets such as derivatives, stocks, currencies, and bonds. Financial markets, which can be virtual or physical, are essential to the world economy. Financial forecasting has become an enormous difficulty due to the markets’ quick internationalization and the volume of data gathered from various sources growing at an accelerating rate. The study explores the creation of a model for stock market forecasting that leverages the power of big data analytics and quantum computing (QC). By handling multifaceted datasets and achieving high-accuracy forecasting, big data and QC improve market trends and insights. This is achieved by leveraging QC speed-driven problem-solving capabilities. In this study, we proposed a novel Hippopotamus Optimized Quantum Refined Support Vector Machine (HO-QRSVM) to predict the financial market. The financial dataset came from several places, such as economic databases and stock exchanges. The preprocessed data from the acquired data were cleaned and normalized. To remove dimensionality and extract features from preprocessed data, principle component analysis (PCA) is used. The capacity of QC to handle entanglement and superposition enables the simultaneous investigation of several potential market situations, leading to faster convergence and more precise forecasts. The result demonstrated the proposed method has significant improvements in prediction accuracy compared to other traditional algorithms. The performance evaluation techniques include the F1-score (96.3%), recall (98.2%), accuracy (98.5%), and precision (98.3%). The research, big data analytics, and QC combined can greatly improve financial market forecasts, giving decision-makers and strategic investors an advantage over their competition.

From Crown Detection to Boundary Segmentation: Advancing Forest Analytics with Enhanced YOLO Model and Airborne LiDAR Point Clouds

Individual tree segmentation is crucial to extract forest structural parameters, which is vital for forest resource management and ecological monitoring. Airborne LiDAR (ALS), with its ability to rapidly and accurately acquire three-dimensional forest structural information, has become an essential tool for large-scale forest monitoring. However, accurately locating individual trees and mapping canopy boundaries continues to be hindered by the overlapping nature of the tree canopies, especially in dense forests. To address these issues, this study introduces CCD-YOLO, a novel deep learning-based network for individual tree segmentation from the ALS point cloud. The proposed approach introduces key architectural enhancements to the YOLO framework, including (1) the integration of cross residual transformer network extended (CReToNeXt) backbone for feature extraction and multi-scale feature fusion, (2) the application of the convolutional block attention module (CBAM) to emphasize tree crown features while suppressing noise, and (3) a dynamic head for adaptive multi-layer feature fusion, enhancing boundary delineation accuracy. The proposed network was trained using a newly generated individual tree segmentation (ITS) dataset collected from a dense forest. A comprehensive evaluation of the experimental results was conducted across varying forest densities, encompassing a variety of both internal and external consistency assessments. The model outperforms the commonly used watershed algorithm and commercial LiDAR 360 software, achieving the highest indices (precision, F1, and recall) in both tree crown detection and boundary segmentation stages. This study highlights the potential of CCD-YOLO as an efficient and scalable solution for addressing the critical challenges of accuracy segmentation in complex forests. In the future, we will focus on enhancing the model’s performance and application.

DPD-YOLO: dense pineapple fruit target detection algorithm in complex environments based on YOLOv8 combined with attention mechanism

With the development of deep learning technology and the widespread application of drones in the agricultural sector, the use of computer vision technology for target detection of pineapples has gradually been recognized as one of the key methods for estimating pineapple yield. When images of pineapple fields are captured by drones, the fruits are often obscured by the pineapple leaf crowns due to their appearance and planting characteristics. Additionally, the background in pineapple fields is relatively complex, and current mainstream target detection algorithms are known to perform poorly in detecting small targets under occlusion conditions in such complex backgrounds. To address these issues, an improved YOLOv8 target detection algorithm, named DPD-YOLO (Dense-Pineapple-Detection YOU Only Look Once), has been proposed for the detection of pineapples in complex environments. The DPD-YOLO model is based on YOLOv8 and introduces the attention mechanism (Coordinate Attention) to enhance the network’s ability to extract features of pineapples in complex backgrounds. Furthermore, the small target detection layer has been fused with BiFPN (Bi-directional Feature Pyramid Network) to strengthen the integration of multi-scale features and enrich the extraction of semantic features. At the same time, the original YOLOv8 detection head has been replaced by the RT-DETR detection head, which incorporates Cross-Attention and Self-Attention mechanisms that improve the model’s detection accuracy. Additionally, Focaler-IoU has been employed to improve CIoU, allowing the network to focus more on small targets. Finally, high-resolution images of the pineapple fields were captured using drones to create a dataset, and extensive experiments were conducted. The results indicate that, compared to existing mainstream target detection models, the proposed DPD-YOLO demonstrated superior detection performance for pineapples in situations where the background is complex and the targets are occluded. The mAP@0.5 reached 62.0%, representing an improvement of 6.6% over the original YOLOv8 algorithm, Precision increased by 2.7%, Recall improved by 13%, and F1-score rose by 10.3%.

Classification of Imagined Speech Signals Using Functional Connectivity Graphs and Machine Learning Models.

EEG involves recording electrical activity generated by the brain through electrodes placed on the scalp. Imagined speech classification has emerged as an essential area of research in brain-computer interfaces (BCIs). Despite significant advances, accurately classifying imagined speech signals remains challenging due to their complex and non-stationary nature. Existing approaches often struggle with low signal-to-noise ratios and high inter-subject variability. A proposed method named imagined speech functional connectivity graph (ISFCG) is implemented to deal with these issues. The functional connectivity graphs capture the complex relationships between brain regions during imagined speech tasks. These graphs are then used to extract features that serve as inputs to various machine-learning models. The ISFCG provides an alternative representation of imagined speech signals, focusing on brain connectivity features to enhance the analysis and classification process. Also, a convolutional neural network (CNN) is proposed to learn features from these complex graphs, leading to improved classification accuracy. Experimental results on a benchmark dataset demonstrate the effectiveness of our method.

A wavelet packet transform-based domain adaptation method for rolling bearings fault diagnosis with fusion of local and global time–frequency features

Abstract Fault diagnosis transfer learning models commonly employ deep neural networks (DNN) to analyze time-frequency features. However, excessively deep neural networks can result in diminished generalization capabilities, leading to subpar performance of the model across various working conditions. Furthermore, inappropriate domain adaptation strategies significantly constrain the accuracy of the model. To address this issue, a Robustly Optimized Residual-Network and Vision Transformer Domain Adaptation Model (RoReViTDAM) is proposed in this article, combining Wavelet Packet Transform (WPT), residual networks, and self-attention mechanisms. Firstly, the WPT is utilized to construct Multi-band Wavelet Coefficient Matrix (MWCM) and corresponding Multi-band Wavelet Coefficient Time-Frequency Feature Matrix (MWSM) with small size and feature aggregation. Subsequently, a shallow Robustly Optimized Residual Network (RoResNet) is designed to effectively extract features from MWCM, considering the spatial distance dependencies of features. Additionally, ViT (Vision Transformer) is employed for time-frequency global feature extraction from MWSM. Furthermore, Domain Adversarial Neural Network (DANN) and Multi-Kernel Maximum Mean Discrepancy (MK-MMD) are employed to extract domain-invariant features from signals of different operating conditions and fault types. At last, three fault diagnosis experiments are conducted in multi-condition scenarios of bearings. The experimental results illustrate the superiority and effectiveness of the proposed model.

Glam Guidance: Empowering Your Fashion Choices with Image Insights

The paper introduces a Fashion Recommendation System that utilizes computer vision and machine learning to analyze fashion items' visual attributes such as color, texture, and style, offering similar or complementary product recommendations to users. It guides readers through the process of building such a system using Python. Leveraging a pre-trained Convolutional Neural Network (CNN), the system extracts feature from fashion images, enabling calculation of similarities between images for recommendation purposes. Employing cosine similarity, it provides personalized recommendations based on user preferences. By integrating advanced image processing techniques, the system offers users a personalized selection of fashion items, enriching their shopping experience and facilitating style exploration. The system's effectiveness is validated through experiments on a diverse dataset of fashion images, demonstrating its capability to deliver accurate and relevant recommendations. If you are interested in developing a Fashion Recommendation System using image features, this article provides a comprehensive guide.

A dual-domain network with division residual connection and feature fusion for CBCT scatter correction.

This study aims to propose a dual-domain network that not only reduces scatter artifacts but also retains structure details in CBCT. The proposed network comprises a projection-domain sub-network and an image-domain sub-network. The projection-domain sub-network utilizes a division residual network to amplify the difference between scatter signals and imaging signals, facilitating the learning of scatter signals. The image-domain sub-network contains dual encoders and a single decoder. The dual encoders extract features from two inputs parallelly, and the decoder fuses the extracted features from the two encoders and maps the fused features back to the final high-quality image. Of the two input images to the image-domain sub-network, one is the scatter-contaminated image analytically reconstructed from the scatter-contaminated projections, and the other is the pre-processed image reconstructed from the pre-processed projections produced by the projection-domain sub-network. Experimental results on both synthetic and real data demonstrate that our method can effectively reduce scatter artifacts and restore image details. Quantitative analysis using synthetic data shows the mean absolute error (MAE) was reduced by 74% and peak signal-to-noise ratio (PSNR) increased by 57% compared to the scatter-contaminated ones. Testing on real data found a 38% increase in contrast-to-noise ratio (CNR) with our method compared to the scatter-contaminated image. Additionally, our method consistently outperforms comparative methods such as U-Net, DSE-Net, RDCNN and the collimator-based method. A dual-domain network that leverages projection-domain division residual connection and image-domain feature fusion has been proposed for CBCT scatter correction. It has potential applications for reducing scatter artifacts and preserving image details in CBCT.