Training Images Research Articles

Cardiovascular health in pediatric patients with X-linked hypophosphatemia under two years of burosumab therapy.

X-linked hypophosphatemia (XLH) is caused by an inactivating mutation in the phosphate-regulating endopeptidase X-linked (PHEX) gene whose defective product fails to control phosphatonin fibroblast growth factor 23 (FGF23) serum levels. Although elevated FGF23 levels have been linked with detrimental cardiac effects, the cardiologic outcomes in XLH patients have been subject to debate. Our study aimed to evaluate the prevalence and severity of cardiovascular morbidity in pediatric XLH patients before, during, and after a 2-year treatment period with burosumab, a recombinant anti-FGF23 antibody. This prospective observational study was conducted in a tertiary medical center, and included 13 individuals with XLH (age range 0.6-16.2 years) who received burosumab every 2 weeks. Clinical assessment at treatment initiation and after .5, 1, and 2 years of uninterrupted treatment included anthropometric measurements and cardiologic evaluations (blood pressure [BP], electrocardiogram, conventional echocardiography, and myocardial strain imaging). The linear growth of all patients improved significantly (mean height z-score: from -1.70 ± 0.80 to -0.96 ± 1.08, P=0.03). Other favorable effects were decline in overweight/obesity rates (from 46.2% to 23.1%) and decreased rates of elevated BP (systolic BP from 38.5% to 15.4%; diastolic BP from 38.5% to 23.1%). Electrocardiograms revealed no significant abnormality throughout the study period. Cardiac dimensions and myocardial strain parameters were within the normative range for age at baseline and remained unchanged during the study period. Cardiologic evaluations provided reassurance that 2 years of burosumab therapy did not cause cardiac morbidity. The beneficial effect of this treatment was a reduction in cardiovascular risk factors, as evidenced by the lower prevalence of both overweight/obesity and elevated BP.

Semi-Supervised Medical Image Classification with Pseudo Labels Using Coalition Similarity Training

The development of medical image classification models necessitates a substantial number of labeled images for model training. In real-world scenarios, sample sizes are typically limited and labeled samples often constitute only a small portion of the dataset. This paper aims to investigate a collaborative similarity learning strategy that optimizes pseudo-labels to enhance model accuracy and expedite its convergence, known as the joint similarity learning framework. By integrating semantic similarity and instance similarity, the pseudo-labels are mutually refined to ensure their quality during initial training. Furthermore, the similarity score is utilized as a weight to guide samples away from misclassification predictions during the classification process. To enhance the model’s generalization ability, an adaptive consistency constraint is introduced into the loss function to improve performance on untrained datasets. The model achieved a satisfactory accuracy of 93.65% at 80% labeling ratio, comparable to supervised learning methods’ performance. Even with very low labeling ratio (e.g., 5%), the model still attained an accuracy of 74.28%. Comparison with other techniques such as Mean Teacher and FixMatch revealed that our approach significantly outperforms them in medical image classification tasks through improving accuracy by approximately 2%, demonstrating this framework’s leadership in medical image classification.

Improving license plate recognition via diverse stylistic plate generation

Deep learning-based license plate recognition heavily relies on the quantity and diversity of training images. However, manually collecting diverse license plate images is a highly time-consuming and labor-intensive task. Hence, generating license plate images to supplement training data emerges as a cost-effective solution for recognition enhancement. Previous generation methods treated the style of license plates as discrete features and extracted style features from a referenced image to generate multi-style license plate images. However, these methods are incapable of generating license plates with arbitrary new styles, as the style of the generated images is constrained by the existing images. To solve this problem, we propose to extract style features from randomly sampled vectors in continuous space for diverse stylistic license plate generation. This way, license plates with arbitrary styles can be generated by simply modifying the sampled style vectors. Besides, to provide controllable content information for the generator, we propose an orientation-aware content encoder to extract multi-oriented glyph content features from the input horizontal glyphs. Moreover, a dual-discriminator is proposed to conduct adversarial learning on both color images and image gradients, aiming to mitigate unnatural patterns and ensure realistic generated results. Finally, extensive experiments on multiple public license plate datasets demonstrate that our method can generate realistic and diverse license plates, and the generated images can significantly improve recognition performance by up to 11.5%.

PW-NeRF: Progressive wavelet-mask guided neural radiance fields view synthesis

Neural Radiance Fields (NeRF) can achieve state-of-the-art new view results when given a sufficient number of training views. However, NeRF's rendering is based on minimizing photometric consistency loss, and during the optimization process, it can suffer from overfitting due to factors such as lighting and texture, resulting in poor geometric and color reconstruction. The less data there is in the overlapping areas of the images, the greater the impact on the results, such as insufficient data caused by occlusion relationships. In this paper, we observed through experiments that introducing low-frequency images into NeRF during training can quickly obtain approximate geometric structures, which can guide NeRF to achieve more stable view synthesis results. Therefore, we propose a progressive wavelet mask to assist in the training of neural radiance fields. By first constructing a gaussian pyramid for the training images and then applying wavelet decomposition to them, we can obtain a series of low-frequency region masks to guide the neural radiance field to focus on learning low-frequency pixel regions and gradually introduce high-frequency lighting and texture changes. Our experiments on the LLFF and Blender datasets show that using a progressive wavelet mask in NeRF training can achieve more realistic generation effects with almost no additional computational overhead. At the same time, we also tested our method in sparse scenes, where it still performs well in avoiding overfitting.

Image inpainting based on GAN-driven structure- and texture-aware learning with application to object removal

In this paper, a novel deep learning-based image inpainting framework consisting of restoring image structure and reconstructing image details from corrupted images is proposed. Most image inpainting methods in the literature aim at restoring image details, outlines, and colors, simultaneously, which may suffer from blurring, deformation, and unreasonable content recovery due to interference among various information. To solve these problems, a two-stage image inpainting deep neural network based on GAN (generative adversarial network) architecture is proposed. The proposed inpainting framework consists of two modules: (1) the first stage, called the structure-aware learning stage, aims at learning a GAN-based structure restoration network, focusing on recovering the low-frequency image component, including colors and outlines of the missing regions of the input corrupted image; and (2) the second stage, called the texture-aware learning stage, aims at learning a GAN-based detail refinement network, focusing on rebuilding the high-frequency image details and texture information. In particular, we also propose to remove details from the training images to better train the structure restoration network to avoid inadequate image structure recovery induced by richer image textures, where the detail reconstruction task is left to the second stage. This strategy achieves to balance the workload between the two stages and the image quality can be progressively enhanced through the two stages. Experimental results have shown that the proposed deep inpainting framework quantitatively and qualitatively achieves state-of-the-art performance on the well-known datasets, including the CelebA, Places2, and ImageNet datasets, compared with existing deep learning-based image inpainting approaches. More specifically, in terms of the two well-known image quality assessment metrics, PSNR (peak signal-to-noise ratio) and SSIM (structural similarity), the improvement percentage of the proposed method, compared with the baseline approach, respectively, ranges from 3.23 % to 11.12 %, and 1.95 % to 13.39 %. The improvements have been shown to stably and significantly outperform the compared state-of-the-art methods in most types of inpainting mask. We also show that the proposed method is applicable to image editing in object removal from a single image.

Multiscale collaborative representation for face recognition via class-information fusion

One of the most challenging issues in face recognition is having only a limited number of training images. Multiscale patch collaborative representation (MSPCRC) is an effective approach to address this problem. However, the existing MSPCRC methods only defined a single patch-scale weight vector for all classes to indicate the importance of different patch scales, ignoring the role of class information when fusing multiscale recognition results. In this work, we consider the effect of class information on face recognition and propose a novel multiscale collaborative-representation face recognition method. Specifically, we first construct the multiscale decision matrices of image subsets from different classes according to patch collaborative representation, and define a patch-scale weight vector for each class to describe the importance of different patch scales. Each element in a scale-weight vector represents the weight value of a certain scale in the corresponding class. Then, we construct the respective optimization objective function for each class, which takes into account the classification information both from the same class and from different classes. Finally, we propose a multiscale fusion-recognition output rule based on the patch-weight vectors. Experimental results demonstrate that the proposed method enhances classification accuracy by approximately 2% to 5% across multiple datasets, surpassing the majority of the competing methods.

Semi-supervised image manipulation localization with residual enhancement

Images have become a significant medium for information transmission, while image forensics has garnered widespread attention from researchers. Due to the scarcity of finely annotated images in the field of image manipulation detection and localization, existing methods for locating manipulated images can only utilize a limited amount of data for model training. However, deep learning models typically require a large number of finely annotated images to fully leverage their powerful fitting capabilities. In this paper, we propose a semi-supervised image manipulation localization framework, employing semi-supervised learning to use unannotated images for deep learning model training. To achieve this goal, we first design a residual enhancement module that contains an encoding-decoding structure. The regression target of this branch is the pristine regions in the images to generate the reconstructed images. Next, we perform a residual operation between the reconstructed and original images to roughly locate the anomalous regions and refine the features extracted by the encoder. Secondly, we employ weakly-supervised learning by adding an image-level classification head to the final layer of the encoder. The classification head generates a class active map based on image-level classification results, which further guides the model’s feature augmentation. Finally, we formulate a specialized semi-supervised framework tailored for image manipulation detection, enabling the utilization of a large volume of unannotated data for model training. Extensive experimental results highlight the notable efficacy of the proposed approach, and outperforms state-of-the-art approaches.

Accuracy of artificial intelligence-assisted endoscopy in the diagnosis of gastric intestinal metaplasia: A systematic review and meta-analysis.

Gastric intestinal metaplasia is a precancerous disease, and a timely diagnosis is essential to delay or halt cancer progression. Artificial intelligence (AI) has found widespread application in the field of disease diagnosis. This study aimed to conduct a comprehensive evaluation of AI's diagnostic accuracy in detecting gastric intestinal metaplasia in endoscopy, compare it to endoscopists' ability, and explore the main factors affecting AI's performance. The study followed the PRISMA-DTA guidelines, and the PubMed, Embase, Web of Science, Cochrane, and IEEE Xplore databases were searched to include relevant studies published by October 2023. We extracted the key features and experimental data of each study and combined the sensitivity and specificity metrics by meta-analysis. We then compared the diagnostic ability of the AI versus the endoscopists using the same test data. Twelve studies with 11,173 patients were included, demonstrating AI models' efficacy in diagnosing gastric intestinal metaplasia. The meta-analysis yielded a pooled sensitivity of 94% (95% confidence interval: 0.92-0.96) and specificity of 93% (95% confidence interval: 0.89-0.95). The combined area under the receiver operating characteristics curve was 0.97. The results of meta-regression and subgroup analysis showed that factors such as study design, endoscopy type, number of training images, and algorithm had a significant effect on the diagnostic performance of AI. The AI exhibited a higher diagnostic capacity than endoscopists (sensitivity: 95% vs. 79%). AI-aided diagnosis of gastric intestinal metaplasia using endoscopy showed high performance and clinical diagnostic value. However, further prospective studies are required to validate these findings.

YOLO-Based Phenotyping of Apple Blotch Disease (Diplocarpon coronariae) in Genetic Resources after Artificial Inoculation

YOLO-Based Phenotyping of Apple Blotch Disease (Diplocarpon coronariae) in Genetic Resources after Artificial Inoculation

Image analysis and teaching strategy optimization of folk dance training based on the deep neural network

To improve the recognition effect of the folk dance image recognition model and put forward new suggestions for teachers’ teaching strategies, this study introduces a Deep Neural Network (DNN) to optimize the folk dance training image recognition model. Moreover, a corresponding teaching strategy optimization scheme is proposed according to the experimental results. Firstly, the image preprocessing and feature extraction of DNN are optimized. Secondly, classification and target detection models are established to analyze the folk dance training images, and the C-dance dataset is used for experiments. Finally, the results are compared with those of the Naive Bayes classifier, K-nearest neighbor, decision tree classifier, support vector machine, and logistic regression models. The results of this study provide new suggestions for teaching strategies. The research results indicate that the optimized classification model shows a significant improvement in classification accuracy across various aspects such as action complexity, dance types, movement speed, dance styles, body dynamics, and rhythm. The accuracy, precision, recall, and F1 scores have increased by approximately 14.7, 11.8, 13.2, and 17.4%, respectively. In the study of factors such as different training images, changes in perspective, lighting conditions, and noise interference, the optimized model demonstrates a substantial enhancement in recognition accuracy and robustness. These findings suggest that, compared to traditional models, the optimized model performs better in identifying various dances and movements, enhancing the accuracy and stability of classification. Based on the experimental results, strategies for optimizing the real-time feedback and assessment mechanism in folk dance teaching, as well as the design of personalized learning paths, are proposed. Therefore, this study holds the potential to be applied in the field of folk dance, promoting the development and innovation of folk dance education.

Computer-aided diagnosis system for optical diagnosis of colorectal polyps under white light imaging

ObjectivesThis study presents a novel computer-aided diagnosis (CADx) designed for optically diagnosing colorectal polyps using white light imaging (WLI).We aimed to evaluate the effectiveness of the CADx and its auxiliary role among endoscopists with different levels of expertise. MethodsWe collected 2,324 neoplastic and 3,735 nonneoplastic polyp WLI images for model training, and 838 colorectal polyp images from 740 patients for model validation. We compared the diagnostic accuracy of the CADx with that of 15 endoscopists under WLI and narrow band imaging (NBI). The auxiliary benefits of CADx for endoscopists of different experience levels and for identifying different types of colorectal polyps was also evaluated. ResultsThe CADx demonstrated an optical diagnostic accuracy of 84.49%, showing considerable superiority over all endoscopists, irrespective of whether WLI or NBI was used (P < 0.001). Assistance from the CADx significantly improved the diagnostic accuracy of the endoscopists from 68.84% to 77.49% (P = 0.001), with the most significant impact observed among novice endoscopists. Notably, novices using CADx-assisted WLI outperform junior and expert endoscopists without such assistance. ConclusionsThe CADx demonstrated a crucial role in substantially enhancing the precision of optical diagnosis for colorectal polyps under WLI and showed the greatest auxiliary benefits for novice endoscopists.

Discriminative diagnosis of ovarian endometriosis cysts and benign mucinous cystadenomas based on the ConvNeXt algorithm

PurposeThe objective of this study was to develop a deep learning model, using the ConvNeXt algorithm, that can effectively differentiate between ovarian endometriosis cysts (OEC) and benign mucinous cystadenomas (MC) by analyzing ultrasound images. The performance of the model in the diagnostic differentiation of these two conditions was also evaluated. MethodsA retrospective analysis was conducted on OEC and MC patients who had sought medical attention at the Fourth Affiliated Hospital of Harbin Medical University between August 2018 and May 2023. The diagnosis was established based on postoperative pathology or the characteristics of aspirated fluid guided by ultrasound, serving as the gold standard. Ultrasound images were collected and subjected to screening and preprocessing procedures. The data set was randomly divided into training, validation, and testing sets in a ratio of 5:3:2. Transfer learning was utilized to determine the initial weights of the ConvNeXt deep learning algorithm, which were further adjusted by retraining the algorithm using the training and validation ultrasound images to establish a new deep learning model. The weights that yielded the highest accuracy were selected to evaluate the diagnostic performance of the model using the validation set. Receiver operating characteristic (ROC) curves were generated, and the area under the curve (AUC) was calculated. Additionally, sensitivity, specificity, positive predictive value, negative predictive value, positive likelihood ratio, negative likelihood ratio, and odds ratio were calculated. Decision curve analysis (DCA) curves were plotted. Results: The study included 786 ultrasound images from 184 patients diagnosed with either OEC or MC. The deep learning model achieved an AUC of 0.90 (95 % CI: 0.85–0.95) in accurately distinguishing between the two conditions, with a sensitivity of 90 % (95 % CI: 84 %-95 %), specificity of 90 % (95 % CI: 77 %-97 %), a positive predictive value of 96 % (95 % CI: 91 %-99 %), a negative predictive value of 77 % (95 % CI: 63 %-88 %), a positive likelihood ratio of 9.27 (95 % CI: 3.65–23.56), and a negative likelihood ratio of 0.11 (95 % CI: 0.06–0.19). The DCA curve demonstrated the practical clinical utility of the model. ConclusionsThe deep learning model developed using the ConvNeXt algorithm exhibits high accuracy (90 %) in distinguishing between OEC and MC. This model demonstrates excellent diagnostic performance and clinical utility, providing a novel approach for the clinical differentiation of these two conditions.

Web-based diagnostic platform for microorganism-induced deterioration on paper-based cultural relics with iterative training from human feedback

PurposePaper-based artifacts hold significant cultural and social values. However, paper is intrinsically fragile to microorganisms, such as mold, due to its cellulose composition, which can serve as a microorganisms’ nutrient source. Mold not only can damage papers’ structural integrity and pose significant challenges to conservation works but also may subject individuals attending the contaminated artifacts to health risks. Current approaches for strain identification usually require extensive training, prolonged time for analysis, expensive operation costs, and higher risks of secondary damage due to sampling. Thus, in current conservation practices with mold-contaminated artifacts, little pre-screening or strain identification was performed before mold removal, and the cleaning techniques are usually broad-spectrum rather than strain-specific. With deep learning showing promising applications across various domains, this study investigated the feasibility of using a convolutional neural network (CNN) for fast in-situ recognition and classification of mold on paper.MethodsMolds were first non-invasively sampled from ancient Xuan Paper-based Chinese books from the Qing and Ming dynasties. Strains were identified using molecular biology methods and the four most prevalent strains were inoculated on Xuan paper to create mockups for image collection. Microscopic images of the molds as well as their stains situated on paper were collected using a compound microscope and commercial microscope lens for cell phone cameras, which were then used for training CNN models with a transfer learning scheme to perform the classification of mold. To enable involvement and contribution from the research community, a web interface that actuates the process while providing interactive features for users to learn about the information of the classified strain was constructed. Moreover, a feedback functionality in the web interface was embedded for catching potential classification errors, adding additional training images, or introducing new strains, all to refine the generalizability and robustness of the model.Results & ConclusionIn the study, we have constructed a suite of high-confidence classification CNN models for the diagnostic process for mold contamination in conservation. At the same time, a web interface was constructed that allows recurrently refining the model with human feedback through engaging the research community. Overall, the proposed framework opens new avenues for effective and timely identification of mold, thus enabling proactive and targeted mold remediation strategies in conservation.

Cardiac screening in pediatric patients with neurofibromatosis type 1: similarities with Noonan syndrome?

Both Neurofibromatosis type 1 (NF1) and Noonan syndrome (NS) are RASopathies. Characteristic cardiac phenotypes of NS, including specific electrocardiographic changes, pulmonary valve stenosis and hypertrophic cardiomyopathy have not been completely studied in NF1. Purpose: The aims of this study were to assess: (1) similarities in the prevalence and types of ECG and conventional echocardiographic findings described in NS in asymptomatic patients with NF1, and (2) the presence of discrete myocardial dysfunction in NF1 patients using myocardial strain imaging. Methods: Fifty-eight patients with NF1 (ages 0–18 years), and thirty-one age-matched healthy controls underwent cardiac assessment including blood pressure measurements, a 12-lead ECG, and detailed echocardiography. Quantification of cardiac chamber size, mass and function were measured using conventional echocardiography. Myocardial strain parameters were assessed using 2-Dimensional (2D) Speckle tracking echocardiography. Results: Asymptomatic patients with NF1 had normal electrocardiograms, none with the typical ECG patterns described in NS. However, patients with NF1 showed significantly decreased calculated Z scores of the left ventricular internal diameter in diastole and systole, reduced left ventricular mass index and a higher incidence of cardiac abnormal findings, mainly of the mitral valve, in contrast to the frequently described types of cardiac abnormalities in NS. Peak and end systolic global circumferential strain were the only significantly reduced speckle tracking derived myocardial strain parameter. Conclusions: Children with NF1 demonstrated more dissimilarities than similarities in the prevalence and types of ECG and conventional echocardiographic findings described in NS. The role of the abnormal myocardial strain parameter needs to be explored.

A methodology for phase characterization in pellet feed using digital microscopy and deep learning

This work proposes a new method for characterizing pellet feed, employing Deep Learning (DL) and Convolutional Neural Networks (CNNs). The main minerals in the composition of the studied pellet feed are hematite, magnetite, goethite, and quartz. Over time, several characterization methodologies have been developed that use Digital Microscopy and Image Analysis tools. The greatest difficulties in this characterization lie in differentiating the textures of hematite particles, the different shapes of their crystals, or discriminating between quartz and resin in reflected light optical microscopy images. This work proposes a mineral characterization methodology based on the Mask R-CNN algorithm. The goal is to perform instance segmentation, that is, to identify, classify, and segment objects in the images. Two DL models were combined: the BF Model performs instance segmentation for the compact, porous, martite, and goethite classes in images obtained in Bright Field mode, and the CPOL Model uses images acquired in Circularly Polarized Light to segment the monocrystalline, polycrystalline, and martite classes. An F1-score of around 80% was obtained for the BF Model and around 90% for the CPOL Model. The results were promising and can be improved as the training image database increases.

Visual Reinforcement Learning for Dynamic Object Detection

Abstract. Object detection is a widely studied task in computer vision. Current methods often focus on images captured from appropriate viewpoints. However, there is a large disparity between objects observed from different viewpoints in the real world. Dynamic Object Detection (DOD) method automatically adjusts the camera viewpoint in a visual scene to sequentially find optimal viewpoints. Currently, the DOD tasks are usually modeled as a sequential decision-making problem and solved using reinforcement learning methods. Existing approaches face challenges with sparse rewards and training instability. To tackle these issues, we proposed a single-step reward function and a lightweight network, respectively. The single-step reward function, which provides timely feedback, gives an efficient training process for DOD tasks. The lightweight network with few parameters can ensure the stability of the training process. To evaluate the effectiveness of our method, we developed a simulation dataset based on UE4, which consists of 1800 training images and 450 testing images. The dataset includes five object categories: vans, cars, trailers, box trucks and SUVs. Experiments demonstrate that our method outperforms SOTA object detectors on our simulation dataset. Specifically, the average precisions(APs) are improved from 89.1% to 96.0% when using the YOLOv8 object detector.

Left atrial strain analysis in the realm of pediatric cardiology: Advantages and implications.

Left atrial (LA) strain analysis has emerged as a noninvasive technique for assessing LA function and early detection of myocardial deformation. Recently, its application has also shown promise in the pediatric population, spanning diverse cardiac conditions that demand accurate and sensitive diagnostic measures. This research article endeavors to explore the role of LA strain parameters and contribute to the growing body of knowledge in pediatric cardiology, paving the way for more effective and tailored approaches to patient care. A comprehensive literature review was conducted to gather evidence from studies using echocardiographic strain imaging techniques across pediatric populations. LA strain parameters exhibited greater sensitivity than conventional atrial function indicators, with early detection of diastolic dysfunction and LA remodeling in pediatric cardiomyopathy, children with multisystem inflammatory syndrome, rheumatic heart disease, as well as childhood renal insufficiency and obesity offering prognostic relevance as potential markers in these pediatric subpopulations. However, there remains a paucity of evidence concerning pediatric mitral valve pathology, justifying further exploration. LA strain analysis carries crucial clinical and prognostic implications in pediatric cardiac conditions, with reliable accuracy and sensitivity to early functional changes.

A novel multi-task machine learning classifier for rare disease patterning using cardiac strain imaging data

To provide accurate predictions, current machine learning-based solutions require large, manually labeled training datasets. We implement persistent homology (PH), a topological tool for studying the pattern of data, to analyze echocardiography-based strain data and differentiate between rare diseases like constrictive pericarditis (CP) and restrictive cardiomyopathy (RCM). Patient population (retrospectively registered) included those presenting with heart failure due to CP (n = 51), RCM (n = 47), and patients without heart failure symptoms (n = 53). Longitudinal, radial, and circumferential strains/strain rates for left ventricular segments were processed into topological feature vectors using Machine learning PH workflow. In differentiating CP and RCM, the PH workflow model had a ROC AUC of 0.94 (Sensitivity = 92%, Specificity = 81%), compared with the GLS model AUC of 0.69 (Sensitivity = 65%, Specificity = 66%). In differentiating between all three conditions, the PH workflow model had an AUC of 0.83 (Sensitivity = 68%, Specificity = 84%), compared with the GLS model AUC of 0.68 (Sensitivity = 52% and Specificity = 76%). By employing persistent homology to differentiate the “pattern” of cardiac deformations, our machine-learning approach provides reasonable accuracy when evaluating small datasets and aids in understanding and visualizing patterns of cardiac imaging data in clinically challenging disease states.

A Generic Model of Traffic Monitoring and Vehicle Counting Systems Based on Edge Computing

One of the main issues in Indonesia is congestion. The number of vehicles continues to increase and is less balanced by the development of transportation infrastructure, especially landlines, causing more complex problems. The Indonesian government needs an intelligent application system that can provide knowledge to unravel congestion. The problem is how to perform edge computing to reduce latency so that the highway monitoring application system runs in real time. This research proposes a basic design for a vehicle monitoring application system that can accurately recognize vehicles, count the number of vehicles, and propose an edge computation that brings computation directly to the data source. The dataset is a video of traffic in Bandung, Jakarta, and several other major cities. The images in the dataset consist of 4,890 training images, 467 validation images, and 231 testing images. In the proposed model, the YOLOv5 and YOLOv7 architectures accurately detect and count vehicles. The test results show a mAP value of 99.1% with an IoU threshold of 50%. Other results include a precision value of 96.2% and a recall of 97.7%. The proposed model can accurately monitor vehicles and reduce latency with an edge computing approach.

Conditional generative model with skip-connection structure for low-light image enhancement

Vision-guided coal mine robots often encounter challenges in low-light environments, capturing images marked by poor visibility and significant loss of detail, which complicates the advancement of smart mine safety production. Despite various advancements, achieving high visibility and detailed images through robotic vision sensors remains a formidable task. This paper introduces a conditional generative model incorporating a skip-connection structure to address these issues. We present a novel approach utilizing conditional Generative Adversarial Networks (GNAs) aimed at improving the visibility of image content. This involves an encoder-decoder framework with skip-connections functioning as the generator, and the discriminator, both tailored to enhance image detail. In addition, we implement loss functions integrated with boundary equilibrium constraints to counter the issue of model collapse in conditional GANs, thereby enhancing training stability and image fidelity. Our experimental findings demonstrate that our model delivers competitive results against other state-of-the-art low-light image enhancement models, alongside a standardized training regimen that ensures rapid and consistent convergence.