Convolutional Neural Network Approach Research Articles

A moment-based pooling approach in convolutional neural networks for breast cancer histopathology image classification

A moment-based pooling approach in convolutional neural networks for breast cancer histopathology image classification

Proto-DS: A Self-Supervised Learning-Based Nondestructive Testing Approach for Food Adulteration with Imbalanced Hyperspectral Data

Conventional food fraud detection using hyperspectral imaging (HSI) relies on the discriminative power of machine learning. However, these approaches often assume a balanced class distribution in an ideal laboratory environment, which is impractical in real-world scenarios with diverse label distributions. This results in suboptimal performance when less frequent classes are overshadowed by the majority class during training. Thus, the critical research challenge emerges of how to develop an effective classifier on a small-scale imbalanced dataset without significant bias from the dominant class. In this paper, we propose a novel nondestructive detection approach, which we call the Dice Loss Improved Self-Supervised Learning-Based Prototypical Network (Proto-DS), designed to address this imbalanced learning challenge. The proposed amalgamation mitigates the label bias on the most frequent class, further improving robustness. We validate our proposed method on three collected hyperspectral food image datasets with varying degrees of data imbalance: Citri Reticulatae Pericarpium (Chenpi), Chinese herbs, and coffee beans. Comparisons with state-of-the-art imbalanced learning techniques, including the Synthetic Minority Oversampling Technique (SMOTE) and class-importance reweighting, reveal our method’s superiority. Notably, our experiments demonstrate that Proto-DS consistently outperforms conventional approaches, achieving the best average balanced accuracy of 88.18% across various training sample sizes, whereas the Logistic Model Tree (LMT), Multi-Layer Perceptron (MLP), and Convolutional Neural Network (CNN) approaches attain only 59.42%, 60.38%, and 66.34%, respectively. Overall, self-supervised learning is key to improving imbalanced learning performance and outperforms related approaches, while both prototypical networks and the Dice loss can further enhance classification performance. Intriguingly, self-supervised learning can provide complementary information to existing imbalanced learning approaches. Combining these approaches may serve as a potential solution for building effective models with limited training data.

A GRAPH CONVOLUTIONAL NETWORK APPROACH FOR PREDICTING NETWORK ROBUSTNESS

Network robustness, which includes controllability robustness and connectivity robustness, reflects the ability of a network system to withstand attacks. In this paper, a Graph Convolutional Network (GCN) approach is proposed for predicting network robustness. In contrast to the existing Convolutional Neural Network (CNN) approach, the network topology and the node characteristics are directly used as GCN input without being converted into a grayscale image. Due to the reduction in the number of feature maps, the model size of a GCN is greatly reduced to only 1% of a CNN. Extensive experimental studies on four representative networks and six real networks have proven that the proposed approach can achieve better predictive performance with less training and running time.

Implementation of Automatic Attendance System Based on Face Recognition Using CNN Method in IPB University Vocational School Environment

The research focuses on creating an automated attendance system using face recognition through the Convolutional Neural Network (CNN) approach at IPB University's Vocational School. The current manual attendance methods show limitations, such as potential inaccuracies in recording and the risk of cheating, like attendance proxies. To overcome these challenges, this study applies the CNN approach with Python and OpenCV, enabling automatic face detection and recognition for students. The system accurately logs attendance by identifying faces in real time. Testing indicates that the system records attendance reliably, whether with a single individual or with multiple faces present within a single frame.

Categorical classification of skin cancer using a weighted ensemble of transfer learning with test time augmentation

Skin cancer is the abnormal development of cells on the surface of the skin and is one of the most fatal diseases in humans. It usually appears in locations that are exposed to the sun, but can also appear in areas that are not regularly exposed to the sun. Due to the striking similarities between benign and malignant lesions, skin cancer detection remains a problem, even for expert dermatologists. Considering the inability of dermatologists to diagnose skin cancer accurately, a convolutional neural network (CNN) approach was used for skin cancer diagnosis. However, the CNN model requires a significant number of image datasets for better performance; thus, image augmentation and transfer learning techniques have been used in this study to boost the number of images and the performance of the model, because there are a limited number of medical images. This study proposes an ensemble transfer-learning-based model that can efficiently classify skin lesions into one of seven categories to aid dermatologists in skin cancer detection: (i) actinic keratoses, (ii) basal cell carcinoma, (iii) benign keratosis, (iv) dermatofibroma, (v) melanocytic nevi, (vi) melanoma, and (vii) vascular skin lesions. Five transfer learning models were used as the basis of the ensemble: MobileNet, EfficientNetV2B2, Xception, ResNext101, and DenseNet201. In addition to the stratified 10-fold cross-validation, the results of each individual model were fused to achieve greater classification accuracy. An annealing learning rate scheduler and test-time augmentation (TTA) were also used to increase the performance of the model during the training and testing stages. A total of 10,015 publicly available dermoscopy images from the HAM10000 (Human Against Machine) dataset, which contained samples from the seven common skin lesion categories, were used to train and evaluate the models. The proposed technique attained 94.49% accuracy on the dataset. These results suggest that this strategy can be useful for improving the accuracy of skin cancer classification. However, the weighted average of f1-score, recall, and precision were obtained to be 94.68%, 94.49%, and 95.07%, respectively.

Automated Recognition of Submerged Body-like Objects in Sonar Images Using Convolutional Neural Networks

The Police Robot for Inspection and Mapping of Underwater Evidence (PRIME) is an uncrewed surface vehicle (USV) currently being developed for underwater search and recovery teams to assist in crime scene investigation. The USV maps underwater scenes using sidescan sonar (SSS). Test exercises use a clothed mannequin lying on the seafloor as a target object to evaluate system performance. A robust, automated method for detecting human body-shaped objects is required to maximise operational functionality. The use of a convolutional neural network (CNN) for automatic target recognition (ATR) is proposed. SSS image data acquired from four different locations during previous missions were used to build a dataset consisting of two classes, i.e., a binary classification problem. The target object class consisted of 166 196 × 196 pixel image snippets of the underwater mannequin, whereas the non-target class consisted of 13,054 examples. Due to the large class imbalance in the dataset, CNN models were trained with six different imbalance ratios. Two different pre-trained models (ResNet-50 and Xception) were compared, and trained via transfer learning. This paper presents results from the CNNs and details the training methods used. Larger datasets are shown to improve CNN performance despite class imbalance, achieving average F1 scores of 97% in image classification. Average F1 scores for target vs background classification with unseen data are only 47% but the end result is enhanced by combining multiple weak classification results in an ensemble average. The combined output, represented as a georeferenced heatmap, accurately indicates the target object location with a high detection confidence and one false positive of low confidence. The CNN approach shows improved object detection performance when compared to the currently used ATR method.

A Deep Learning Approach to Identify Potential Sites for Pocket Park Installation in Nanjing, China

ABSTRACT Pocket parks are considered valuable tools for supporting green space provision in high-density environments. As the factors affecting pocket parks vary in fine-grained granularity, analyzing them efficiently on a large scale using traditional methods is challenging. Therefore, considering Nanjing as the study site, a potential identification framework was established, which includes categories of space, vitality, facility, and pleasurability for pocket park installation (PPI) based on visual information from street view images (SVIs). Then, the TrueSkill algorithm was employed to construct four parallel-task training sets according to the four categories and a convolutional neural network (CNN) and active learning (AL) approaches were applied to establish four parallel-task models of deep learning (DL). A two-phase grading score method was designed to integrate the identification results of the four DL models into a comprehensive potential score to determine priority sites for PPI. Lastly, the attributes of the potential sites and corresponding planning strategies in different scenarios were analyzed. The results showed that this modelling approach can integrally assess multiple factors on PPI within an enormous scope, overcoming traditional methods’ spatial and dimensional constraint in high-density environments. Our method and results can inform future pocket park planning and design in Nanjing and similar cities worldwide.

Tree sequences as a general-purpose tool for population genetic inference.

As population genetics data increases in size new methods have been developed to store genetic information in efficient ways, such as tree sequences. These data structures are computationally and storage efficient, but are not interchangeable with existing data structures used for many population genetic inference methodologies such as the use of convolutional neural networks (CNNs) applied to population genetic alignments. To better utilize these new data structures we propose and implement a graph convolutional network (GCN) to directly learn from tree sequence topology and node data, allowing for the use of neural network applications without an intermediate step of converting tree sequences to population genetic alignment format. We then compare our approach to standard CNN approaches on a set of previously defined benchmarking tasks including recombination rate estimation, positive selection detection, introgression detection, and demographic model parameter inference. We show that tree sequences can be directly learned from using a GCN approach and can be used to perform well on these common population genetics inference tasks with accuracies roughly matching or even exceeding that of a CNN-based method. As tree sequences become more widely used in population genetics research we foresee developments and optimizations of this work to provide a foundation for population genetics inference moving forward.

Evaluating Convolutional Neural Network Architecture for Historical Topographic Hardcopy Maps Analysis: A Study on Training and Validation Accuracy Variation

Convolutional Neural Networks (CNN) are widely used for image analysis tasks, including object detection, segmentation, and recognition. Given the advanced capability, this study evaluates the effectiveness and performance of CNN architecture for analysing Historical Topographic Hardcopy Maps (HTHM) by assessing variations in training and validation accuracy. The lack of research specifically dedicated to CNN’s application in analysing topographic hardcopy maps presents an opportunity to explore and address the unique challenges associated with this domain. While existing studies have predominantly focused on satellite imagery, this study aims to uncover valuable insights, patterns, and characteristics inherent to HTHM through customised CNN approaches. This study utilises a standard CNN architecture and tests the model’s performance with different epoch settings (20, 40, and 60) using varying dataset sizes (288, 636, 1144, and 1716 images). The results indicate that the optimal operation point for training and validation accuracy is achieved at epoch 40. Beyond epoch 40, the widening gap between training and validation accuracy suggests overfitting. Hence, adding more epochs does not significantly improve accuracy beyond the optimum phase. The experiment also shows that the CNN model obtains a training accuracy of 98%, validation accuracy of 67%, and F1-score overall performance of 77%. The analysis demonstrates that the CNN model performs reasonably well in classifying instances from the HTHM dataset. These findings contribute to a better understanding of the strengths and limitations of the model, providing valuable insights for future research and refinement of classification approaches in the context of topographic hardcopy map analysis.

Enhancing Accuracy of Flame Equivalence Ratio Measurements: An Attention-Based Convolutional Neural Network Approach for Overcoming Limitations in Traditional Color Modeling.

This paper addresses the inherent limitations in traditional color modeling techniques for measuring the flame equivalence ratio (Φ), particularly focusing on the subjectivity involved in threshold settings and the challenges posed by uneven 2D color distribution. To overcome these issues, this study introduces an attention-based convolutional neural network (ACN) model, a novel approach that transcends the conventional reliance on B/G color features (Tf). The ACN model leverages adaptive feature extraction, augmented by a spatial attention mechanism, to more effectively analyze flame images. By amplifying key features, autonomously minimizing background noise, and standardizing variations in color distribution, the ACN model in this experiment achieved a prediction accuracy of 99%, with a 76% reduction in error rate compared to the original model, significantly improving the accuracy and objectivity of flame Φ measurement. This method marks a substantial development in the precision and reliability of flame analysis.

Point-enhanced convolutional neural network: A novel deep learning method for transonic wall-bounded flows

Low order models can be used to accelerate engineering design processes. Ideally, these surrogates should meet the conflicting requirements of large design space coverage, high accuracy and fast evaluation. Within the context of aerospace applications at transonic conditions, this can be challenging due to the associated non-linearity of the flow regime. Different methods have been investigated in the past to predict the flow-field around shapes such as airfoils or cylinders. However, they usually have reduced spatial resolution, limiting the prediction capabilities within the boundary layer which is of interest for transonic wall-bounded flows. This work proposes a novel Point-Enhanced Convolutional Neural Network (PCNN) method that combines the advantages of the well-established PointNet and convolutional neural network approaches. The PCNN model has relatively low memory requirements in the training process, preserves the spatial correlation in the domain and has the same resolution as a traditional computational method. The architecture is used for the flow-field prediction of civil aero-engine nacelles in which it is demonstrated that the flow features of peak isentropic Mach number (Mis), pre-shock isentropic Mach number and shock location (X/Lnac) are captured within ΔMis = 0.02, ΔMis=0.04, ΔX/Lnac=0.007, respectively. The PCNN model successfully predicts the integral parameters of the boundary layer, in which the incompressible displacement thickness, momentum thickness and shape factor are typically within 5% of the CFD. Overall, the PCNN method is demonstrated for transonic wall-bounded flows for a range of flow physics that include shock waves and shock-induced separation.

Optimizing Fractional-Order Convolutional Neural Networks for Groove Classification in Music Using Differential Evolution

This study presents a differential evolution (DE)-based optimization approach for fractional-order convolutional neural networks (FOCNNs) aimed at enhancing the accuracy of groove classification in music. Groove, an essential element in music perception, is typically influenced by rhythmic patterns and acoustic features. While FOCNNs offer a promising method for capturing these subtleties through fractional-order derivatives, they face challenges in efficiently converging to optimal parameters. To address this, DE is applied to optimize the initial weights and biases of FOCNNs, leveraging its robustness and ability to explore a broad solution space. The proposed DE-FOCNN was evaluated on the Janata dataset, which includes pre-rated music tracks. Comparative experiments across various fractional-order values demonstrated that DE-FOCNN achieved superior performance in terms of higher test accuracy and reduced overfitting compared to a standard FOCNN. Specifically, DE-FOCNN showed optimal performance at fractional-order values such as v = 1.4. Further experiments demonstrated that DE-FOCNN achieved higher accuracy and lower variance compared to other popular evolutionary algorithms. This research primarily contributes to the optimization of FOCNNs by introducing a novel DE-based approach for the automated analysis and classification of musical grooves. The DE-FOCNN framework holds promise for addressing other related engineering challenges.

A Novel Detection Transformer Framework for Ship Detection in Synthetic Aperture Radar Imagery Using Advanced Feature Fusion and Polarimetric Techniques

Ship detection in synthetic aperture radar (SAR) imagery faces significant challenges due to the limitations of traditional methods, such as convolutional neural network (CNN) and anchor-based matching approaches, which struggle with accurately detecting smaller targets as well as adapting to varying environmental conditions. These methods, relying on either intensity values or single-target characteristics, often fail to enhance the signal-to-clutter ratio (SCR) and are prone to false detections due to environmental factors. To address these issues, a novel framework is introduced that leverages the detection transformer (DETR) model along with advanced feature fusion techniques to enhance ship detection. This feature enhancement DETR (FEDETR) module manages clutter and improves feature extraction through preprocessing techniques such as filtering, denoising, and applying maximum and median pooling with various kernel sizes. Furthermore, it combines metrics like the line spread function (LSF), peak signal-to-noise ratio (PSNR), and F1 score to predict optimal pooling configurations and thus enhance edge sharpness, image fidelity, and detection accuracy. Complementing this, the weighted feature fusion (WFF) module integrates polarimetric SAR (PolSAR) methods such as Pauli decomposition, coherence matrix analysis, and feature volume and helix scattering (Fvh) components decomposition, along with FEDETR attention maps, to provide detailed radar scattering insights that enhance ship response characterization. Finally, by integrating wave polarization properties, the ability to distinguish and characterize targets is augmented, thereby improving SCR and facilitating the detection of weakly scattered targets in SAR imagery. Overall, this new framework significantly boosts DETR’s performance, offering a robust solution for maritime surveillance and security.

Optimizing secure multimedia communication in embedded systems a parallel convolutional neural network approach to RIS and D2D resource allocation

With the rapid development of Internet of Things (IoT) services, technologies that leverage multimedia computer communication for information sharing in embedded systems have become a research focus. To address the challenges of low spectral efficiency and poor network flexibility in multimedia computer communications, this paper proposes a resource allocation scheme based on parallel Convolutional Neural Network (CNN). The scheme optimizes the base station beamforming vector and the Reconfigurable Intelligent Surface (RIS) phase shifts to maximize the secure transmission rate for cellular users (CUs), while ensuring normal and secure communication for device-to-device (D2D) users. First, to mitigate interference caused by D2D users reusing CU spectrum resources, the RIS phase shifts and beamforming vectors are optimized to suppress interference and enhance system secrecy rates. Second, to maximize the CU secrecy rate, the paper proposes a parallel CNN-based resource allocation model that considers base station transmission power, RIS reflection coefficients, and D2D communication rate constraints, incorporating multi-scale residual modules in the convolutional layers of the model. Simulation results demonstrate that the proposed CNN-based resource allocation scheme significantly improves the secrecy rate of embedded system communications, ensuring secure multimedia computing, and outperforms traditional methods.

Deep Learning-Enabled Dynamic Model for Nutrient Status Detection of Aquaponically Grown Plants

Developing models to assess the nutrient status of plants at various growth stages is challenging due to the dynamic nature of plant development. Hence, this study encoded spatiotemporal information of plants within a single time-series model to precisely assess the nutrient status of aquaponically cultivated lettuce. In particular, the long short-term memory (LSTM) and deep autoencoder (DAE) approaches were combined to classify aquaponically grown lettuce plants according to their nutrient status. The proposed approach was validated using extensive sequential hyperspectral reflectance measurements acquired from lettuce leaves at different growth stages across the growing season. A DAE was used to extract distinct features from each sequential spectral dataset time step. These features were used as input to an LSTM model to classify lettuce grown across a gradient of nutrient levels. The results demonstrated that the LSTM outperformed the convolutional neural network (CNN) and multi-class support vector machine (MCSVM) approaches. Also, features selected by the DAE showed better performance compared to features extracted using both genetic algorithms (GAs) and sequential forward selection (SFS). The hybridization of deep autoencoder and long short-term memory (DAE-LSTM) obtained the highest overall classification accuracy of 94%. The suggested methodology presents a pathway to automating the process of nutrient status diagnosis throughout the entire plant life cycle, with the LSTM technique poised to assume a pivotal role in forthcoming time-series analyses for precision agriculture.

Tree sequences as a general-purpose tool for population genetic inference.

As population genetics data increases in size new methods have been developed to store genetic information in efficient ways, such as tree sequences. These data structures are computationally and storage efficient, but are not interchangeable with existing data structures used for many population genetic inference methodologies such as the use of convolutional neural networks (CNNs) applied to population genetic alignments. To better utilize these new data structures we propose and implement a graph convolutional network (GCN) to directly learn from tree sequence topology and node data, allowing for the use of neural network applications without an intermediate step of converting tree sequences to population genetic alignment format. We then compare our approach to standard CNN approaches on a set of previously defined benchmarking tasks including recombination rate estimation, positive selection detection, introgression detection, and demographic model parameter inference. We show that tree sequences can be directly learned from using a GCN approach and can be used to perform well on these common population genetics inference tasks with accuracies roughly matching or even exceeding that of a CNN-based method. As tree sequences become more widely used in population genetics research we foresee developments and optimizations of this work to provide a foundation for population genetics inference moving forward.

A scalable convolutional neural network approach to fluid flow prediction in complex environments

We evaluate the capability of convolutional neural networks (CNNs) to predict a velocity field as it relates to fluid flow around various arrangements of obstacles within a two-dimensional, rectangular channel. We base our network architecture on a gated residual U-Net template and train it on velocity fields generated from computational fluid dynamics (CFD) simulations. We then assess the extent to which our model can accurately and efficiently predict steady flows in terms of velocity fields associated with inlet speeds and obstacle configurations not included in our training set. Real-world applications often require fluid-flow predictions in larger and more complex domains that contain more obstacles than used in model training. To address this problem, we propose a method that decomposes a domain into subdomains for which our model can individually and accurately predict the fluid flow, after which we apply smoothness and continuity constraints to reconstruct velocity fields across the whole of the original domain. This piecewise, semicontinuous approach is computationally more efficient than the alternative, which involves generation of CFD datasets required to retrain the model on larger and more spatially complex domains. We introduce a local orientational vector field entropy (LOVE) metric, which quantifies a decorrelation scale for velocity fields in geometric domains with one or more obstacles, and use it to devise a strategy for decomposing complex domains into weakly interacting subsets suitable for application of our modeling approach. We end with an assessment of error propagation across modeled domains of increasing size.

Convolutional neural network approach for fault detection and characterization in medium voltage distribution networks

Power outages significantly impact the power industry by disrupting social welfare and economic stability. Still, existing methods for fault detection face challenges due to load and network topology, conditions, and installed equipment. However, recent advances in artificial intelligence (AI) are enabling researchers to create alternative approaches for fault detection and location strategies. Therefore, this paper introduces a novel method for detecting, classifying, and locating faults in power systems through voltage waveform analysis using a convolutional neural network (CNN) integrated with the Piecewise Function Put Together (PFPT) algorithm for fault detection and fault zone localization in a power distribution network. Utilizing Park's transformation, noise reduction PFPT sine fitting, and CNNs, the proposed method distinguishes between 'healthy' and 'faulty' conditions. Simulation results reveal that while the voltage Park's vector time behavior of a healthy system remains stable, it exhibits circular or mixed patterns under faulty conditions. These patterns enable the identification of four types of short circuit faults—single-line-to-ground (LG), line-to-line (LL), line-to-line-to-ground (LLG), and three-line (3L) faults—by analyzing 3D voltage Park's waveforms at network buses. The study validates fault type identification through the observation of rotating Park vectors from sine fitting of time-based voltage waveforms. By converting 3D voltage waveforms into high-resolution images, the method utilizes a CNN for fault recognition, achieving an accuracy of 93.1%. This innovative approach underscores the robustness and precision of combining traditional electrical engineering techniques with modern AI.

Advanced hybrid empirical mode decomposition, convolutional neural network and long short-term memory neural network approach for predicting grain pile humidity based on meteorological inputs

Grain pile humidity prediction is beneficial to ensure food security, and establishing an effective humidity prediction model is of great significance to the field of grain storage. By taking meteorological and grain temperature data as inputs, we propose a prediction model that combines Empirical Mode Decomposition (EMD), Convolutional Neural Network (CNN), and Long Short-Term Memory Network (LSTM). The model was verified in experimental data of three different storage layers of grain piles. The prediction results show that the proposed EMD-CNN-LSTM model has better prediction accuracy than the other three comparison models: CNN-LSTM, CNN and LSTM. From the average results of the entire granary, the MAE, RMSE, and MAPE results are 0.14, 0.18, and 0.25%, respectively, and the MAE value is 44% higher than the previous research method that does not consider meteorological factors. The MAE, RMSE, and MAPE results of the CNN-LSTM method with EMD decomposition were improved by 58%, 53% and 58% respectively compared with the method without EMD decomposition. It can be concluded that taking meteorological factors as model input and integrating EMD methods can improve prediction accuracy. The constructed prediction model shows effective prediction results in different storage layers of grain pile, which provides new insights for ensuring food security and also provides valuable references for multivariate time series prediction in other fields.

Improving Crowd Counting Performance: A Convolutional Neural Network Approach with Transfer Learning

Improving Crowd Counting Performance: A Convolutional Neural Network Approach with Transfer Learning