Deep Convolutional Network Model Research Articles

Data augmentation-based enhanced fingerprint recognition using deep convolutional generative adversarial network and diffusion models

The progress of fingerprint recognition applications encounters substantial hurdles due to privacy and security concerns, leading to limited fingerprint data availability and stringent data quality requirements. This article endeavors to tackle the challenges of data scarcity and data quality in fingerprint recognition by implementing data augmentation techniques. Specifically, this research employed two state-of-the-art generative models in the domain of deep learning, namely Deep Convolutional Generative Adversarial Network (DCGAN) and the Diffusion model, for fingerprint data augmentation. Generative Adversarial Network (GAN), as a popular generative model, effectively captures the features of sample images and learns the diversity of the sample images, thereby generating realistic and diverse images. DCGAN, as a variant model of traditional GAN, inherits the advantages of GAN while alleviating issues such as blurry images and mode collapse, resulting in improved performance. On the other hand, Diffusion, as one of the most popular generative models in recent years, exhibits outstanding image generation capabilities and surpasses traditional GAN in some image generation tasks. The experimental results demonstrate that both DCGAN and Diffusion can generate clear, high-quality fingerprint images, fulfilling the requirements of fingerprint data augmentation. Furthermore, through the comparison between DCGAN and Diffusion, it is concluded that the quality of fingerprint images generated by DCGAN is superior to the results of Diffusion, and DCGAN exhibits higher efficiency in both training and generating images compared to Diffusion.

Research on Parkinson's disease recognition algorithm based on sample enhancement

Parkinson's disease patients have early vocal cord damage, and their voiceprint characteristics differ significantly from those of healthy individuals, which can be used to identify Parkinson's disease. However, the samples of the voiceprint dataset of Parkinson's disease patients are insufficient, so this paper proposes a double self-attention deep convolutional generative adversarial network model for sample enhancement to generate high-resolution spectrograms, based on which deep learning is used to recognize Parkinson's disease. This model improves the texture clarity of samples by increasing network depth and combining gradient penalty and spectral normalization techniques, and a family of pure convolutional neural networks (ConvNeXt) classification network based on Transfer learning is constructed to extract voiceprint features and classify them, which improves the accuracy of Parkinson's disease recognition. The validation experiments of the effectiveness of this paper's algorithm are carried out on the Parkinson's disease speech dataset. Compared with the pre-sample enhancement, the clarity of the samples generated by the proposed model in this paper as well as the Fréchet inception distance (FID) are improved, and the network model in this paper is able to achieve an accuracy of 98.8%. The results of this paper show that the Parkinson's disease recognition algorithm based on double self-attention deep convolutional generative adversarial network sample enhancement can accurately distinguish between healthy individuals and Parkinson's disease patients, which helps to solve the problem of insufficient samples for early recognition of voiceprint data in Parkinson's disease. In summary, the method effectively improves the classification accuracy of small-sample Parkinson's disease speech dataset and provides an effective solution idea for early Parkinson's disease speech diagnosis.

Performance Improvement of Deep Convolutional Networks for Aerial Imagery Segmentation of Natural Disaster-Affected Areas

This study proposes a framework for improving performance and exploring the application of Deep Convolutional Networks (DCN) using the best parameters and criteria to accurately produce aerial imagery semantic segmentation of natural disaster-affected areas. This study utilizes two models: U-Net and Pyramid Scene Parsing Network (PSPNet). Extensive study results show that the Grid Search algorithm can improve the performance of the two models used, whereas previous research has not used the Grid Search algorithm to improve performance in aerial imagery segmentation of natural disaster-affected areas. The Grid Search algorithm performs parameter tuning on DCN, data augmentation criteria tuning, and dataset criteria tuning for pre-training. The most optimal DCN model is shown by PSPNet (152) (bpc), using the best parameters and criteria, with a mean Intersection over Union (mIoU) of 83.34%, a significant mIoU increase of 43.09% compared to using only the default parameters and criteria (baselines). The validation results using the k-fold cross-validation method on the most optimal DCN model produced an average accuracy of 99.04%. PSPNet(152) (bpc) can detect and identify various objects with irregular shapes and sizes, can detect and identify various important objects affected by natural disasters such as flooded buildings and roads, and can detect and identify objects with small shapes such as vehicles and pools, which are the most challenging task for semantic segmentation network models. This study also shows that increasing the network layers in the PSPNet-(18, 34, 50, 101, 152) model, which uses the best parameters and criteria, improves the model's performance. The results of this study indicate the need to utilize a special dataset from aerial imagery originating from the Unmanned Aerial Vehicle (UAV) during the pre-training stage for transfer learning to improve DCN performance for further research.

An Improved MobileNetV3 Mushroom Quality Classification Model Using Images with Complex Backgrounds

Shiitake mushrooms are an important edible fungus, and their nutrient content is related to their quality. With the acceleration of urbanization, there has been a serious loss of population and shortage of labor in rural areas. The problem of harvesting agricultural products after maturity is becoming more and more prominent. In recent years, deep learning techniques have performed well in classification tasks using image data. These techniques can replace the manual labor needed to classify the quality of shiitake mushrooms quickly and accurately. Therefore, in this paper, a MobileNetV3_large deep convolutional network is improved, and a mushroom quality classification model using images with complex backgrounds is proposed. First, captured image data of shiitake mushrooms are divided into three categories based on the appearance characteristics related to shiitake quality. By constructing a hybrid data set, the model’s focus on shiitake mushrooms in images with complex backgrounds is improved. And the constructed data set is expanded using data enhancement methods to improve the generalization ability of the model. The total number of images after expansion is 10,991. Among them, the number of primary mushroom images is 3758, the number of secondary mushroom images is 3678, and the number of tertiary mushroom images is 3555. Subsequently, the SE module in MobileNetV3_large network is improved and processed to enhance the model recognition accuracy while reducing the network size. Finally, PolyFocalLoss and migration learning strategies are introduced to train the model and accelerate model convergence. In this paper, the recognition performance of the improved MobileNetV3_large model is evaluated by using the confusion matrix evaluation tool. It is also compared with other deep convolutional network models such as VGG16, GoogLeNet, ResNet50, MobileNet, ShuffleNet, and EfficientNet using the same experimental conditions. The results show that the improved MobileNetV3_large network has a recognition accuracy of 99.91%, a model size of 11.9 M, and a recognition error rate of 0.09% by the above methods. Compared to the original model, the recognition accuracy of the improved model is increased by 18.81% and the size is reduced by 26.54%. The improved MobileNetV3_large network model in this paper has better comprehensive performance, and it can provide a reference for the development of quality recognition and classification technologies for shiitake mushrooms cultivated in greenhouse environments.

AI-based evaluation system for supply chain vulnerabilities and resilience amidst external shocks: An empirical approach

The study focuses on the intricacies and vulnerabilities inherent in supply chains, which are often influenced by external disruptions such as pandemics, conflict scenarios, and inflation. The aim is to devise an AI-driven system that can accurately appraise these intricacies within the domain and mitigate their vulnerabilities effectively. The work employs an empirical approach utilizing datasets from various studies for developing Machine Learning (ML) and Deep Learning (DL) models. These include linear regression, deep learning, CNN networks are designed to predict supply chain risks and enhance the overall stability and performance of an industrial supply chain system. The Deep CNN regression model outperforms as compared to the other models in predicting supply chain risks, achieving an accuracy rate of approximately 90%. The developed model will be more proficient in dealing with complex and nonlinear relationships among the variables. The study introduces a novel approach to data augmentation using Fuzzy C-means in conjunction with a Deep Convolution network model. This approach expands the data size, reduces the forecast errors, and reduces the computational of the given model. The results highlight the potential of ML and DL in enhancing predictability and resilience in the face of escalating risks within supply chain networks. The findings of the work will significantly offer insights for strategists and planners in the supply chain sector to enhance their operational effectiveness and efficiency.

Automatic Face Recognition of Target Images Based on Deep Learning Algorithms

The current research results of face recognition have poor recognition accuracy, time-consuming and high information loss rate. A method of automatic face recognition based on depth learning algorithm is proposed. The three primary RGB images are transformed into color saturated HSV images, and the dynamic range is extended by changing the non-linearity according to the conversion results. Based on the extended situation, the membership function is used to map the target image to the blurred plane, and the contrast enhancement of the image is completed. The enhanced image is substituted into the image segmentation, and the image eigenvalues are extracted preliminarily. The seed of image segmentation is generated by the eigenvalues. The growing seed regions are merged by the color distance and the texture distance to achieve the target image segmentation. Based on image segmentation, LBP operator is used to extract the local texture features of face twice, and then a deep convolution network model is constructed. The shared weights of convolution network model and pooling down-sampling technique are used to reduce the complexity of the model. The top layer of the model forms a feature classification surface of face image, fuses the constraint conditions, and obtains the trained face. The deep convolution network model is used to extract features from face images and complete face recognition. The experimental results show that the method has good accuracy, high efficiency and low information loss rate.

A deep graph convolutional network model of NOx emission prediction for coal‐fired boiler

AbstractTo deal with environmental problems caused by NOx production in thermal plants, it is imperative to establish a reliable model to predict NOx concentration in the combustion process. NOx formation in a coal‐fired boiler is complex, and many variables affect NOx emissions. The effective information fusion of these variables can improve the accuracy of NOx concentration prediction. However, the existing NOx prediction algorithms based on thermal parameters rarely consider the mechanical knowledge of the boiler operation, and it is not easy to incorporate the topological information of production into modelling. Therefore, a graph convolutional network is proposed for NOx emission prediction. First, the key variables affecting NOx generation are selected according to the knowledge and the random forest‐based variable importance. Then, the model structure is designed by exploring the topological information among thermal variables to capture the complex spatial dependence. The model inputs are constructed by coding different operation variables, and the adjacency matrix is generated according to the correlation information between variables, which can fuse data information and reduce redundancy. On this basis, the prediction model of NOx concentration is established. Historical data from a 660 MW coal‐fired boiler are used in the experiment. The prediction results show that the proposed model can effectively fuse the information of characteristic variables and fully exploit the non‐linear mapping relationship between process variables and NOx emission. When compared with three typical models in NOx modelling, the proposed model has better performance with a determination coefficient of 0.906.

Marine chlorophyll-a prediction based on deep auto-encoded temporal convolutional network model

Chlorophyll-a concentration is one of the important indicators for assessing marine ecology. Accurate prediction of marine chlorophyll-a concentrations is a prerequisite for early warning of marine hazards such as red tide. There are many factors affecting chlorophyll-a concentration. Marine water quality parameter data as a time series is non-linear. There is a strong coupling relationship between various water quality parameters. This makes the prediction difficult. Therefore, there is a need to conduct research on marine chlorophyll-a concentration prediction methods. In this study, a depth-automatically encoded temporal convolutional network (DAE-TCN) is proposed to predict marine chlorophyll-a concentrations. Specifically, the DAE-TCN algorithm performs spatial feature extraction and data dimensionality reduction on the data through encoding to alleviate the difficulty of temporal feature extraction by temporal convolution, improve the efficiency of data information utilization, and obtain accurate prediction results. In this paper, we use the data from the coast of Beihai, Guangxi, as the experimental data. The DAE-TCN algorithm is compared with three baseline algorithms, and the results indicate that the algorithm shows advantages in stability and prediction accuracy and can be well applied to marine chlorophyll-a concentration prediction. This study has scientific significance for marine environmental protection.

Infrared Dim and Small Target Sequence Dataset Generation Method Based on Generative Adversarial Networks

With the development of infrared technology, infrared dim and small target detection plays a vital role in precision guidance applications. To address the problems of insufficient dataset coverage and huge actual shooting costs in infrared dim and small target detection methods, this paper proposes a method for generating infrared dim and small target sequence datasets based on generative adversarial networks (GANs). Specifically, first, the improved deep convolutional generative adversarial network (DCGAN) model is used to generate clear images of the infrared sky background. Then, target–background sequence images are constructed using multi-scale feature extraction and improved conditional generative adversarial networks. This method fully considers the infrared characteristics of the target and the background, which can achieve effective expansion of the image data and provide a test set for the infrared small target detection and recognition algorithm. In addition, the classifier’s performance can be improved by expanding the training set, which enhances the accuracy and effect of infrared dim and small target detection based on deep learning. After experimental evaluation, the dataset generated by this method is similar to the real infrared dataset, and the model detection accuracy can be improved after training with the latest deep learning model.

SM-TCNNET: A High-Performance Method for Detecting Human Activity Using WiFi Signals

Human activity recognition (HAR) is an important research area with a wide range of application scenarios, such as smart homes, healthcare, abnormal behavior detection, etc. Wearable sensors, computer vision, radar, and other technologies are commonly used to detect human activity. However, they are severely limited by issues such as cost, lighting, context, and privacy. Therefore, this paper explores a high-performance method of using channel state information (CSI) to identify human activities, which is a deep learning-based spatial module-temporal convolutional network (SM-TCNNET) model. The model consists of a spatial feature extraction module and a temporal convolutional network (TCN) that can extract the spatiotemporal features in CSI signals well. In this paper, extensive experiments are conducted on the self-picked dataset and the public dataset (StanWiFi), and the results show that the accuracy reaches 99.93% and 99.80%, respectively. Compared with the existing methods, the recognition accuracy of the SM-TCNNET model proposed in this paper is improved by 1.8%.

A Tourist Attractions Recommendation Model Based on Sequence Interest Extraction and Location Information Fusion

Travel, as one way to relax oneself, has become the first choice for people to enjoy their body and mind in modern society. However, while facing lots of information, how to help users make better decisions on their next travel goals through their historical interest spots is a direction that needs further research in big data recommendation systems. In this thesis, we proposed the deep convolution and multi-head self-attention position network model. First, it extracts the user’s historical interest point feature information by convolutional neural network method, and then performs horizontal and vertical filtering. Next, it interacts the obtained information with the candidate attraction information, and extracts the location information of the historical interest sequence by the multi-head self-attention mechanism. Finally, the model does the attention mechanism of the candidate attraction by fusing the feature information of the location information. The final model achieves a deep fusion of user sequence interest and location feature information. We conducted detailed comparison experiments with the very popular models in the industry on different public datasets, and the results showed that our deep convolution and multi-head self-attention position network model has good performance.

Optimization of W-DCGAN Fault Diagnosis Method Based on Self-attention Mechanism

Bearing component is the most important component of wind turbine, the harsh operating environment makes the bearing prone to failure, and the maintenance cost is very expensive when a failure occurs. Complex environmental noise brings serious noise pollution to fault samples. In addition, the serious sample imbalance between fault samples and normal samples brings great challenges to bearing fault diagnosis. In view of the above mentioned problems, this paper uses a self-attention mechanism optimization and Wasserstein distance improvement deep convolutional adversarial network model based on self-attention mechanism optimization and Wasserstein distance improvement deep Bearing fault diagnosis method based on the convolution generative Adversarial Network Model (SAW-DCGAN). The experimental results show that W-DCGAN has excellent generating ability and can generate samples like real samples to achieve the balance of samples. The addition of self-attention makes the classification features more expressive, and can accelerate the training speed of the model while having a higher fault diagnosis rate, which verifies the practicability of the method.

Bone age assessment based on deep learning architecture

<span lang="EN-US">The fast advancement of technology has prompted the creation of automated systems in a variety of sectors, including medicine. One application is an automated bone age evaluation from left-hand X-ray pictures, which assists radiologists and pediatricians in making decisions about the growth status of youngsters. However, one of the most difficult aspects of establishing an automated system is selecting the best approach for producing effective and dependable predictions, especially when working with large amounts of data. As part of this work, we investigate the use of the convolutional neural networks (CNNs) model to classify the age of the bone. The work’s dataset is based on the radiological society of North America (RSNA) dataset. To address this issue, we developed and tested deep learning architecture for autonomous bone assessment, we design a new deep convolution network (DCNN) model. The assessment measures that use in this work are accuracy, recall, precision, and F-score. The proposed model achieves 97% test accuracy for bone age classification.</span>

Assessing uncertainty and heterogeneity in machine learning-based spatiotemporal ozone prediction in Beijing-Tianjin- Hebei region in China

Accurate prediction of spatiotemporal ozone concentration is of great significance to effectively establish advanced early warning systems and regulate air pollution control. However, the comprehensive assessment of uncertainty and heterogeneity in spatiotemporal ozone prediction remains unknown. Here, we systematically analyze the hourly and daily spatiotemporal predictive performances using convolutional long short term memory (ConvLSTM) and deep convolutional generative adversarial network (DCGAN) models over the Beijing-Tianjin-Hebei region in China from 2013 to 2018. In extensive scenarios, our results show that the machine learning-based (ML-based) models achieve better spatiotemporal ozone concentration prediction performance with multiple meteorological conditions. A further comparison to the air pollution model-Nested Air Quality Prediction Modelling System (NAQPMS) and monitoring observations, the ConvLSTM model demonstrates the practical feasibility of identifying high ozone concentration distribution and capturing spatiotemporal ozone variation patterns at a high spatial resolution (here 15 km × 15 km).

Underwater Acoustic Target Recognition Based on Data Augmentation and Residual CNN

In the field of underwater acoustic recognition, machine learning methods rely on a large number of datasets to achieve high accuracy, while the actual collected signal samples are often very scarce, which has a great impact on the recognition performance. This paper presents a recognition method of an underwater acoustic target by the data augmentation technique and the residual convolutional neural network (CNN) model, which is used to expand training samples to improve recognition performance. As a representative model in residual CNN, the ResNet18 model is used for recognition. The whole process mainly includes mel-frequency cepstral coefficient (MFCC) feature extraction, data augmentation processing, and ResNet18 model recognition. On the base of the traditional data augmentation, this study used the deep convolutional generative adversarial network (DCGAN) model to realize the expansion of underwater acoustic samples and compared the recognition performance of support vector machine (SVM), common CNN, VGG19, and ResNet18. The recognition results of the MFCC, constant Q transform (CQT), and low-frequency analyzer and recorder (LOFAR) spectrum were also analyzed and compared. Experimental results showed that the recognition accuracy of the MFCC feature was better than that of other features at the same method, and using the data augmentation method could obviously improve the recognition performance. Moreover, the recognition performance of ResNet18 using data enhancement technology was better than that of other models, which was due to the combination of the data expansion advantage of data augmentation technology and the deep feature extracting ability of the residual CNN model. In addition, although this method was used for ship recognition in this paper, it is not limited to this. This method is also applicable to other target voice recognition, such as natural sound and underwater voice biometrics.

HRRNet: Hierarchical Refinement Residual Network for Semantic Segmentation of Remote Sensing Images

Semantic segmentation of high-resolution remote sensing images plays an important role in many practical applications, including precision agriculture and natural disaster assessment. With the emergence of a large number of studies on convolutional neural networks, the performance of the semantic segmentation model of remote sensing images has been dramatically promoted. However, many deep convolutional network models do not fully refine the segmentation result maps, and, in addition, the contextual dependencies of the semantic feature map have not been adequately exploited. This article proposes a hierarchical refinement residual network (HRRNet) to address these issues. The HRRNet mainly consists of ResNet50 as the backbone, attention blocks, and decoders. The attention block consists of a channel attention module (CAM) and a pooling residual attention module (PRAM) and residual structures. Specifically, the feature map output by the four blocks of Resnet50 is passed through the attention block to fully explore the contextual dependencies of the position and channel of the semantic feature map, and, then, the feature maps of each branch are fused step by step to realize the refinement of the feature maps, thereby improving the segmentation performance of the proposed HRRNet. Experiments show that the proposed HRRNet improves segmentation result maps compared with various state-of-the-art networks on Vaihingen and Potsdam datasets.

BSNet: Boundary-semantic-fusion network for farmland parcel mapping in high-resolution satellite images

Mapping farmland parcels using satellite images is essential for agricultural remote sensing applications. Loss of spatial details and positioning of parcel boundaries are the main challenges in available deep convolution network (DCN) models. This study developed a boundary-semantic-fusion DCN (BSNet) model for delineating farmland parcels from high-resolution satellite images. Central to this method is the combination between shallow-level boundary features with accurate spatial positioning and deep-level semantic features for category identification. First, a general deep convolution framework consisting of boundary, semantic and fusion blocks was implemented for farmland parcel mapping. Second, a particular serial structure with a detaching operation for linking the boundary and semantic blocks was explored to maintain the spatial details and fine-scale boundaries in feature learning. Third, an encoder-decoder fusion block was developed to integrate the boundary and semantic features to produce the final parcel maps. We validated the proposed model with different high-resolution satellite images in two study areas. The experimental results, with improvements greater than 4% in the F1 score and 6% in the IoU score relative to other comparative methods, illustrate the effectiveness of the proposed model for fine-scale farmland parcel mapping.

MODCN: Fine-Tuned Deep Convolutional Neural Network with GAN Deployed to Forecast Diabetic Eye Damage in Fundus Retinal Images

Diabetic Retinopathy (DR) and Glaucoma are two of the most common causes of vision loss world-wide. However, it can be averted if therapy is begun early enough. In biomedical applications, the use of digital image processing has assisted in the automated identification of some ailments at an earlier stage. To make this prediction generally neural network classifier models were previously used, but these models have the drawback of being unable to detect multiple illnesses that occur in the eye at the same time and require a big database for successful classifier training. As a result, a model is needed to reliably distinguish DR and Glaucoma in diabetic individuals more accurately and with minimum dataset images. In this view, this study introduced Mayfly Optimized Deep Convolutional Network (MODCN) model for automated disease detection in the fundus retina images. In the MODCN model, the images are initially preprocessed, segmented at generator in the GAN model then a discriminator readily gives synthesis of real images of the fundus retina, thus a wide database has been created and considered as training images for the MODCN classifier. MODCN classifier has a modified high-density layer as a transition layer to avoid overfitting and the errors are minimized by tuning the hyperparameters using Mayfly Optimization Algorithm. After feature mapping, the classes normal, DR and Glaucoma are labeled and stored. At the testing stage, images are preprocessed, feature mapped and classified in the MODCN model. Thus, the proposed MODCN model detects multiple illness such as Diabetic Retinopathy and Glaucoma at the same time even with a small amount of database that performs a successful classifier training. This model is then evaluated and gives an accuracy of 99% that was higher compared to previous models.

State evaluation of copper flotation process based on transfer learning and a layered and blocked framework

AbstractState evaluation is vital to ensure the process operating optimality for copper flotation processes. Specifically, the froth image is the comprehensive embodiment of raw ore properties and process operations, which is one of the key factors to realize condition recognition and state evaluation. Firstly, a feature mosaic technique‐based neural network framework is proposed. The input image features are extracted from the different network structures, which can achieve higher precision in condition recognition and state evaluation than a single neural network framework. Then, an improved deep convolutional generative adversarial networks (DCGAN) model based on feature matching and maximize mean discrepancy (MMD) distance is investigated so that the froth images with high similarity, integrity, and balance to the original images can be generated. Therefore, the problem of small image sets and the lack of labelled images for some sub‐processes can be solved. Finally, a layered and blocked state evaluation model is constructed based on the improved DCGAN model and transfer learning (TL) so that the state evaluation of the copper flotation process with multiple sub‐processes, long process, and small image sets of some sub‐processes is solved. The effectiveness of the proposed method is verified through a series of data experiments on a copper flotation industrial process.

HTTP Traffic Analysis based on Multiple Deep Convolution Network Model Generation Algorithms

In recent years, with the development of the Internet, social networking, online banking, e-commerce and other network applications are growing rapidly. At the same time, all kinds of malicious web pages are constantly emerging. Under the new situation, the network security threats are distributed, large-scale and complex. New network attack modes are emerging. With more and more diverse devices access to the Internet, our life is more intelligent and convenient, but also brings more loopholes and hidden dangers. Some malicious web pages through a variety of means to lure users to open URL links and conduct malicious behavior. However, if we can detect the URL of the malicious web page and identify the malicious web page, we can avoid the problems of content variability and behavior tracking. Therefore, traffic analysis based on various deep convolution network model generation algorithms arises at the historic moment, and becomes an important research issue in the field of Internet security.