CNN Network Research Articles

Image-based rice leaf disease detection using CNN and generative adversarial network

Image-based rice leaf disease detection using CNN and generative adversarial network

SSA: semantic structure aware inference on CNN networks for weakly pixel-wise dense predictions without cost

SSA: semantic structure aware inference on CNN networks for weakly pixel-wise dense predictions without cost

A compressed video quality enhancement algorithm based on CNN and transformer hybrid network

A compressed video quality enhancement algorithm based on CNN and transformer hybrid network

An optimized model for network intrusion detection in the network operating system environment

With the heavy reliance on computers and information technology to send and receive data across networks of various types, there has been concern about securing that data from intrusions and cyber-attacks. The expansion of network usage has led to an increase in hacker attacks, which has led to prioritizing cybersecurity precautions in detecting potential threats. Intrusion detection techniques are a critical security measure to protect networks in both personal and corporate environments that are managed by network operating systems. For this, the paper relies on designing a network intrusion detection model. Since deep neural networks (DNNs) are classic deep learning models known for their strong classification performance, making them popular in intrusion detection along with other machine learning algorithms, they have been chosen to improve intrusion classification models based on datasets for intrusion detection systems. The basic structure of this proposal is to adopt one of the optimization algorithms in extracting features from the dataset to obtain more accurate results in the classification and intrusion detection stage. The developed Corona Virus algorithm is adopted to improve the system performance by identifying optimal features. This algorithm, which consists of several stages, optimally selects individuals based on features from the NSL-KDD dataset used for intrusion detection. The resulting optimization solution acts as a network structure for the intrusion classification model based on machine learning and deep learning algorithms. The test results showed exceptional performance on the NSL-KDD dataset, where the proposed Convolution Neural Network CNN model achieved 99.3% accuracy for multi-class classification, while the Decision Tree (DT) achieved 88.64% accuracy for anomaly detection in bi-class classification.

PII-GCNet: Lightweight Multi-Modal CNN Network for Efficient Crowd Counting and Localization in UAV RGB-T Images

With the increasing need for real-time crowd evaluation in military surveillance, public safety, and event crowd management, crowd counting using unmanned aerial vehicle (UAV) captured images has emerged as an essential research topic. While conventional RGB-based methods have achieved significant success, their performance is severely hampered in low-light environments due to poor visibility. Integrating thermal infrared (TIR) images can address this issue, but existing RGB-T crowd counting networks, which employ multi-stream architectures, tend to introduce computational redundancy and excessive parameters, rendering them impractical for UAV applications constrained by limited onboard resources. To overcome these challenges, this research introduces an innovative, compact RGB-T framework designed to minimize redundant feature processing and improve multi-modal representation. The proposed approach introduces a Partial Information Interaction Convolution (PIIConv) module to selectively minimize redundant feature computations and a Global Collaborative Fusion (GCFusion) module to improve multi-modal feature representation through spatial attention mechanisms. Empirical findings indicate that the introduced network attains competitive results on the DroneRGBT dataset while significantly reducing floating-point operations (FLOPs) and improving inference speed across various computing platforms. This study’s significance is in providing a computationally efficient framework for RGB-T crowd counting that balances accuracy and resource efficiency, making it ideal for real-time UAV deployment.

Coronary Artery Disease Detection Based on a Novel Multi-Modal Deep-Coding Method Using ECG and PCG Signals.

Coronary artery disease (CAD) is an irreversible and fatal disease. It necessitates timely and precise diagnosis to slow CAD progression. Electrocardiogram (ECG) and phonocardiogram (PCG), conveying abundant disease-related information, are prevalent clinical techniques for early CAD diagnosis. Nevertheless, most previous methods have relied on single-modal data, restricting their diagnosis precision due to suffering from information shortages. To address this issue and capture adequate information, the development of a multi-modal method becomes imperative. In this study, a novel multi-modal learning method is proposed to integrate both ECG and PCG for CAD detection. Along with deconvolution operation, a novel ECG-PCG coupling signal is evaluated initially to enrich the diagnosis information. After constructing a modified recurrence plot, we build a parallel CNN network to encode multi-modal information, involving ECG, PCG and ECG-PCG coupling deep-coding features. To remove irrelevant information while preserving discriminative features, we add an autoencoder network to compress feature dimension. Final CAD classification is conducted by combining support vector machine and optimal multi-modal features. The experiment is validated on 199 simultaneously recorded ECG and PCG signals from non-CAD and CAD subjects, and achieves high performance with accuracy, sensitivity, specificity and f1-score of 98.49%, 98.57%,98.57% and 98.89%, respectively. The result demonstrates the superiority of the proposed multi-modal method in overcoming information shortages of single-modal signals and outperforming existing models in CAD detection. This study highlights the potential of multi-modal deep-coding information, and offers a wider insight to enhance CAD diagnosis.

A multi-step regularity assessment and joint prediction system for ordering time series based on entropy and deep learning

Customer maintenance is of vital importance to the enterprise management. Valuable assessment and efficient prediction for customer ordering behavior can offer better decision-making and reduce business costs significantly. According to existing studies about customer behavior regularity segment and demand prediction most focus on e-commerce and other fields with large amount of data, making them not suitable for small enterprises and data features like sparsity and outliers are not mined when doing regularity quantification. Additionally, more and more complex network structures for demand prediction are proposed, which builds on the assumption that all the samples have predictive value, ignoring the fine-grained analysis of different time series regularity with high cost. To deal with the above issues, a multi-step regularity assessment and joint prediction system for ordering time series is proposed. For extracting features, comprehensive assessment of customer regularity based on entropy weight method with the result of predictability quantification using K-Means clustering algorithm, real entropy, LZW algorithm and anomaly detection adopting Isolation Forest algorithm not only gives an objective result to ‘how high the regularity of customers is’, filling the gap in the field of regularity quantification, but also provides a theoretical basis for demand prediction models selection. Prediction models: Random Forest regression, XGBoost, CNN and LSTM network are experimented with sMAPE and MSLE for performance evaluation to verify the effectiveness of the proposed regularity quantitation method. Moreover, a merged CNN-BiLSTM neural network model is established for predicting those customers with low regularity and difficult to predict by traditional machine leaning algorithms, which performs better on the data set compared to others. Random Forest is still used for prediction of customers with high regularity due to its high training efficiency. Finally, the results of prediction, regularity quantification, and classification are output from the intelligent system, which is capable of providing scientific basis for corporate strategy decision and has highly extendibility in other enterprises and fields for follow-up research.

PBNet: Combining Transformer and CNN in Passport Background Texture Printing Image Classification

Passport background texture classification has always been an important task in border checks. Current manual methods struggle to achieve satisfactory results in terms of consistency and stability for weakly textured background images. For this reason, this study designs and develops a CNN and Transformer complementary network (PBNet) for passport background texture image classification. We first design two encoders by Transformer and CNN to produce complementary features in the Transformer and CNN domains, respectively. Then, we cross-wisely concatenate these complementary features to propose a feature enhancement module (FEM) for effectively blending them. In addition, we introduce focal loss to relieve the overfitting problem caused by data imbalance. Experimental results show that our PBNet significantly surpasses the state-of-the-art image segmentation models based on CNNs, Transformers, and even Transformer and CNN combined models designed for passport background texture image classification.

Interactive transformer and CNN network for fusion classification of hyperspectral and LiDAR data

ABSTRACT The Transformer has become pivotal for the integrated analysis of multi-source remote-sensing (RS) data in Earth observation, particularly in applications such as the fusion classification of hyperspectral images (HSI) and Light Detection and Ranging (LiDAR) data. However, Transformers are often employed as effective feature extractors by adopting similar processing blocks for different modalities from multi-source sensors, overlooking differences in imaging principles and data characteristics. Moreover, in the feature extraction process across different sensor data, there is a lack of necessary cross-modal information interaction, leading to insufficient utilization of complementary information between different sensors and resulting in suboptimal fusion outcomes. In this paper, we propose an interactive Transformer and CNN network for the fusion classification of HSI and LiDAR data. Specifically, a heterogeneous three-branch network architecture is designed for HSI and LiDAR data, where Transformers and CNNs encapsulate global contextual spatial and spectral information for HSI and capture geometric elevation patterns for LiDAR data, respectively. Elevation-Spatial Interaction (ESI) and Spectral-Spatial Interaction (SSI) modules are then introduced for multi-stage feature interaction. ESI enables the CNN-Transformer network to focus on essential local elevation details while simultaneously modelling global contextual spatial information. SSI facilitates the Transformer-Transformer network to cyclically intertwine spectral and spatial information for long-range spectral-spatial feature fusion. Finally, the interacted elevation, spatial, and spectral features undergo the Gated Fusion module to achieve hierarchical fusion adaptively, resulting in an elevation-spatial-spectral representation. Experiments conducted on three benchmark HSI-LiDAR datasets demonstrate the effectiveness of our proposed approach.

Recognizing human activities using light-weight and effective machine learning methodologies

Background Human activity recognition poses a complex challenge in predicting individuals’ movements from raw sensor data using machine learning models. This paper explores the application of six prominent machine learning techniques – decision tree, random forest, linear regression, Naïve Bayes, k-nearest neighbor, and neural networks – to enhance the accuracy of human activity detection for e-health systems. Despite previous research efforts employing data mining and machine learning, there remains room for improvement in performance. The study focuses on predicting activities such as walking, standing, laying, sitting, walking upstairs, and walking downstairs. Methods The research employs six machine learning algorithms to recognize human activities, including decision tree, random forest, linear regression, Naïve Bayes, k-nearest neighbor, and neural networks. Results Evaluation of the human activity recognition dataset reveals that the random forest classifier, CNN, GRN and neural network yield promising results, achieving high accuracy. However, Naïve Bayes falls short of satisfying outcomes. Conclusions The study successfully classifies activities like SITTING, STANDING, LAYING, WALKING, WALKING_DOWNSTAIRS, and WALKING_UPSTAIRS with a remarkable accuracy of 98%. The contribution lies in the thorough exploration of machine learning techniques, with neural networks emerging as the most effective in enhancing human activity recognition. The findings showcase the potential for advanced applications in e-health systems and beyond.

Lightweight Hotspot Detection Model Fusing SE and ECA Mechanisms.

In this paper, we propose a lightweight lithography machine learning-based hotspot detection model that integrates the Squeeze-and-Excitation (SE) attention mechanism and the Efficient Channel Attention (ECA) mechanism. These mechanisms can adaptively adjust channel weights, significantly enhancing the model's ability to extract relevant features of hotspots and non-hotspots through cross-channel interaction without dimensionality reduction. Our model extracts feature vectors through seven convolutional layers and four pooling layers, followed by three fully connected layers that map to the output, thereby simplifying the CNN network structure. Experimental results on our collected layout dataset and the ICCAD 2012 layout dataset demonstrate that our model is more lightweight. By evaluating overall accuracy, recall, and runtime, the comprehensive performance of our model is shown to exceed that of ConvNeXt, Swin transformer, and ResNet 50.

Deep learning for automated encrustation detection in sewer inspection

Rapid urbanization and population growth in recent decades have placed significant pressure on urban cities to rely heavily on underground infrastructure, such as sewers and tunnels, to maintain the provision of essential services. These sewers, typically having a limited lifespan of 50 to 100 years, are prone to various forms of defects. While prior research has primarily addressed common sewer defect like crack, root intrusion, and infiltration among others, the challenge of encrustation—the formation of hard deposits within sewer systems—has received less attention. This study presents a pioneering deep-learning approach to detect encrustation in sewers by leveraging survey videos from 14 different sewers in the United Kingdom. Our work marks the first effort to develop models specifically for detecting encrustation using deep learning techniques, as previous studies have focused on other types of deposits such as settled and attached deposits. By converting the videos into sequential image frames, we subjected them to thorough analysis and several image pre-processing techniques. Our contributions include the development and comparison of different classification models using backbone CNN networks such as AlexNet, VGG16, EfficientNet, and VGG19 to classify encrustation. Notably, this study provides the first metric-based comparison of these backbone networks to identify the most effective model for encrustation detection. The results demonstrate an impressive 96 % accuracy using the deep architecture of VGG19. Beyond accuracy, this research explores the impact of data augmentation and network dropout on reducing overfitting and enhancing model performance. Additionally, we analyze the time complexities associated with training models with and without data augmentation, providing valuable insights into the efficiency of our approach.

A paralleled CNN and Transformer network for PPG-based cuff-less blood pressure estimation

Cuff-less blood pressure measurement plays a pivotal part in hypertension diagnosis and prevention. By employing deep learning algorithms, blood pressure estimation applying photoplethysmography (PPG) signals has achieved remarkable improvements and attracted growing attention. However, most existing methods utilize the PPG signal emphasized on localized feature learning, often neglecting the global features in the temporal data. To tackle the issue, we propose a novel architecture. It leverages a parallel processing approach by using Convolutional Neural Network for extracting local features and simultaneously employing Transformer to capture global features. The feature fusion block integrates the features using spatial and channel attention. The blood pressure values are regressed by two fully connected layers. To verify the performance of our method, we conduct evaluations on the Medical Information Mart for Intensive Care (MIMIC) database. On the MIMIC database, our algorithm demonstrates exceptional performance. For Diastolic Pressure, Mean Arterial Pressure, and Systolic Pressure, the mean absolute errors are 2.36 mmHg, 2.03 mmHg, and 4.44 mmHg, respectively. This performance aligns with the rigorous criteria required by an Association for the Advancement of Medical Instrumentation (AAMI) and receives a Grade A rating in accordance with the British Hypertension Society (BHS) standard.

Non-contrasted computed tomography (NCCT) based chronic thromboembolic pulmonary hypertension (CTEPH) automatic diagnosis using cascaded network with multiple instance learning

Objective.The diagnosis of chronic thromboembolic pulmonary hypertension (CTEPH) is challenging due to nonspecific early symptoms, complex diagnostic processes, and small lesion sizes. This study aims to develop an automatic diagnosis method for CTEPH using non-contrasted computed tomography (NCCT) scans, enabling automated diagnosis without precise lesion annotation.Approach.A novel cascade network (CN) with multiple instance learning (CNMIL) framework was developed to improve the diagnosis of CTEPH. This method uses a CN architecture combining two Resnet-18 CNN networks to progressively distinguish between normal and CTEPH cases. Multiple instance learning (MIL) is employed to treat each 3D CT case as a 'bag' of image slices, using attention scoring to identify the most important slices. An attention module helps the model focus on diagnostically relevant regions within each slice. The dataset comprised NCCT scans from 300 subjects, including 117 males and 183 females, with an average age of 52.5 ± 20.9 years, consisting of 132 normal cases and 168 cases of lung diseases, including 88 cases of CTEPH. The CNMIL framework was evaluated using sensitivity, specificity, and the area under the curve (AUC) metrics, and compared with common 3D supervised classification networks and existing CTEPH automatic diagnosis networks.Main results. The CNMIL framework demonstrated high diagnostic performance, achieving an AUC of 0.807, accuracy of 0.833, sensitivity of 0.795, and specificity of 0.849 in distinguishing CTEPH cases. Ablation studies revealed that integrating MIL and the CN significantly enhanced performance, with the model achieving an AUC of 0.978 and perfect sensitivity (1.000) in normal classification. Comparisons with other 3D network architectures confirmed that the integrated model outperformed others, achieving the highest AUC of 0.8419.Significance. The CNMIL network requires no additional scans or annotations, relying solely on NCCT. This approach can improve timely and accurate CTEPH detection, resulting in better patient outcomes.

A Review of ECG Classification Techniques Based FPGA

Electrocardiogram (ECG) signals are particularly used in the diagnosis and supervision of different cardiovascular diseases. The accurate and efficient classification of ECG signals is of critical importance in clinical applications. FPGA-based systems have been identified to offer an efficient solution to the implementation of real-time ECG signal classification systems because of their parallel processing, reprogrammable nature, and low power consumption. Thus, this review paper aims to explore the existing ECG classification techniques with a focus on FPGA implementation. The review discussed several classification methods, including adaptive algorithms and neural networks, which have been optimized for arrhythmia detection. Notably, Neural network techniques like ANN, CNN, and spiking neural networks are preferable due to their ability to provide superior accuracy. For instance, Mathias et al. developed a high-precision neural network with an impressive accuracy of 99.82%. Similarly, Arona et al. deployed a DCNN on FPGA, achieving accuracies reaching up to 99.67%. Researchers widely used the MIT-BIH dataset to evaluate ECG classification techniques, This paper presents new trends and focuses on future research prospects in this field. Hence, the findings of the present review can help researchers and engineers in developing effective and efficient FPGA-based ECG monitoring and diagnostic systems for clinical and healthcare applications.

HAD-Net: An attention U-based network with hyper-scale shifted aggregating and max-diagonal sampling for medical image segmentation

Objectives:Accurate extraction of regions of interest (ROI) with variable shapes and scales is one of the primary challenges in medical image segmentation. Current U-based networks mostly aggregate multi-stage encoding outputs as an improved multi-scale skip connection. Although this design has been proven to provide scale diversity and contextual integrity, there remain several intuitive limits: (i) the encoding outputs are resampled to the same size simply, which destruct the fine-grained information. The advantages of utilization of multiple scales are insufficient. (ii) Certain redundant information proportional to the feature dimension size is introduced and causes multi-stage interference. And (iii) the precision of information delivery relies on the up-sampling and down-sampling layers, but guidance on maintaining consistency in feature locations and trends between them is lacking. Methods:To improve these situations, this paper proposed a U-based CNN network named HAD-Net, by assembling a new hyper-scale shifted aggregating module (HSAM) paradigm and progressive reusing attention (PRA) for skip connections, as well as employing a novel pair of dual-branch parameter-free sampling layers, i.e. max-diagonal pooling (MDP) and max-diagonal un-pooling (MDUP). That is, the aggregating scheme additionally combines five subregions with certain offsets in the shallower stage. Since the lower scale-down ratios of subregions enrich scales and fine-grain context. Then, the attention scheme contains a partial-to-global channel attention (PGCA) and a multi-scale reusing spatial attention (MRSA), it builds reusing connections internally and adjusts the focus on more useful dimensions. Finally, MDP and MDUP are explored in pairs to improve texture delivery and feature consistency, enhancing information retention and avoiding positional confusion. Results:Compared to state-of-the-art networks, HAD-Net has achieved comparable and even better performances with Dice of 90.13%, 81.51%, and 75.43% for each class on BraTS20, 89.59% Dice and 98.56% AUC on Kvasir-SEG, as well as 82.17% Dice and 98.05% AUC on DRIVE. Conclusions:The scheme of HSAM+PRA+MDP+MDUP has been proven to be a remarkable improvement and leaves room for further research.

Privacy-SF: An encoding-based privacy-preserving segmentation framework for medical images

Deep learning is becoming increasingly popular and is being extensively used in the field of medical image analysis. However, the privacy sensitivity of medical data limits the availability of data, which constrains the advancement of medical image analysis and impedes collaboration across multiple centers. To address this problem, we propose a novel encoding-based framework, named Privacy-SF, aimed at implementing privacy-preserving segmentation for medical images. Our proposed segmentation framework consists of three CNN networks: 1) two encoding networks on the client side that encode medical images and their corresponding segmentation masks individually to remove the privacy features, 2) a unique mapping network that analyzes the content of encoded data and learns the mapping from the encoded image to the encoded mask. By sequentially encoding data and optimizing the mapping network, our approach ensures privacy protection for images and masks during both the training and inference phases of medical image analysis. Additionally, to further improve the segmentation performance, we carefully design augmentation strategies specifically for encoded data based on its sequence nature. Extensive experiments conducted on five datasets with different modalities demonstrate excellent performance in privacy-preserving segmentation and multi-center collaboration. Furthermore, the analysis of encoded data and the experiment of model inversion attacks validate the privacy-preserving capability of our approach.

A rotated object detection strategy for remote sensing images using misaligned cross-fusion structures

Remote sensing image object detection is a challenging task with a focus on detection accuracy. In the object detection task of remote sensing images, scenes often contain objects with varying scales and dense arrays, leading to issues such as low accuracy, missed detections, and false alarms. To address these challenges, this paper introduces MCSC-Net, a rotated object detector, designed to enhance detection performance for objects of different scales in densely arranged scenarios. The proposed detector features a novel neck network named MCFN-V3, which incorporated a cross-fusion structure. MCFN-V3 facilitates simultaneous feature extraction in upper and lower layers with feedback to the middle layer. Through the cross-fusion of feature layers, it achieves the integration of multi-scale feature information, thereby enhancing the network's detection capabilities for objects at multiple scales. Additionally, the paper presents the STC module to address the problem of poor feature correlation in high-resolution image processing by CNN networks. This module improves feature expressiveness and range, strengthening inter-feature relationships within layers. Furthermore, in the target positioning stage, the paper transforms the angle regression problem into a classification task, making the network more suitable for detecting rotated objects in dense scenes. To assess the effectiveness of the algorithm, experiments were conducted on two publicly available remote sensing datasets, DOTA and UCAS-AOD. Our method achieved a multiclass average precision (mAP) of 77.4 % on DOTA and an outstanding performance of 97.0 % on UCAS-AOD. These experimental results validate the effectiveness of the proposed approach.

A robust adapted Flexible Parallel Neural Network architecture for early prediction of lithium battery lifespan

Early prediction of End of Life (EPEOL) is crucial for improving lithium battery efficiency and lifespan. Traditional fixed-architecture neural networks often suffer from underfitting or overfitting due to diverse data distributions. To address this, we propose the Flexible Parallel Neural Network (FPNN). This model integrates modules like InceptionBlock, 3D CNN, 2D CNN, and dual-stream networks. By effectively extracting electrochemical features from video-format data through the 3D CNN and achieving multi-scale feature abstraction with InceptionBlock, FPNN ensures effective module coordination. The model can adaptively adjust the number of InceptionBlocks to handle tasks of varying complexity. Experimental results on the MIT dataset show that FPNN achieves MAPE values of 1.26%, 0.41%, 0.37%, 0.33%, 0.32%, 0.32%, 0.31%, 0.31%, 0.22%, and 0.34% using the first 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100 cycle data, respectively. The interpretability of FPNN is reflected in its structural design and flexible unit choices, providing a basis for model interpretation. Comprising multiple modules, FPNN enables smooth feature extraction from electrochemical data through collaborative interaction. The diverse branching structure of the flexible units allows the model to capture features at different scales, learning richer information. Our approach offers a precise, adaptable, and easy-to-understand solution for EPEOL, opening new possibilities in battery health monitoring.

A Study of Electric Bicycle Lithium Battery Charging Monitoring Using CNN and BiLSTM Networks Model with NILM Method

A Study of Electric Bicycle Lithium Battery Charging Monitoring Using CNN and BiLSTM Networks Model with NILM Method