Convolutional Neural Network Design Research Articles

Inverse design of convolutional neural networks via nanophotonic kernels

We proposed an ultra-compact neural network based on inversely designed photonic kernels, which has the advantages of a small footprint, high computing speed, and low power consumption. The core layers of the neural network, composed of automatically optimized photonic multiplication kernels and addition kernels, can perform arbitrary photonic computations at the speed of the light wave, offering more advantages than their electronic counterparts. Furthermore, the convolution, pooling, and fully-connected layers can be realized by the appropriate permutation of the proposed optical kernels. We constructed an all-optical convolutional neural network with a high accuracy of 97.7% for recognizing handwritten digits. Our results could significantly promote the inverse design of photonic neural networks.

Vision-aided grasp classification: design and evaluation of compact CNN for prosthetic hands

Powered prosthetic hands capable of executing various grasp patterns are highly sought-after solutions for upper limb amputees. A crucial requirement for such prosthetic hands is the accurate identification of the intended grasp pattern and subsequent activation of the prosthetic digits accordingly. Vision-based grasp classification techniques offer improved coordination between amputees and prosthetic hands without physical contact. Deep learning methods, particularly Convolutional Neural Networks (CNNs), are utilized to process visual information for classification. The key challenge lies in developing a model that can effectively generalize across various object shapes and accurately classify grasp classes. To address this, a compact CNN model named GraspCNet is proposed, specifically designed for grasp classification in prosthetic hands. The use of separable convolutions reduces the computational burden, making it potentially suitable for real-time applications on embedded systems. The GraspCNet model is designed to learn and generalize from object shapes, allowing it to effectively classify unseen objects beyond those included in the training dataset. The proposed model was trained and tested using various standard object data sets. A cross-validation strategy has been adopted to perform better in seen and unseen object class scenarios. The average accuracy achieved was 82.22% and 75.48% in the case of seen, and unseen object classes respectively. In computer-based real-time experiments, the GraspCNet model achieved an accuracy of 69%. A comparative analysis with state-of-the-art techniques revealed that the proposed GraspCNet model outperformed most benchmark techniques and demonstrated comparable performance with the DcnnGrasp method. The compact nature of the GraspCNet model suggests its potential for integration with other sensing modalities in prosthetic hands.

Automatic multi-label diagnosis of single-lead ECG using novel hybrid residual recurrent convolutional neural networks

Arrhythmia is a prevalent cardiovascular disease that can unveil various heart health issues. In recent times, the rise of wearable devices has garnered attention towards the advantages of portable single-lead ECG devices in heart health monitoring. Achieving end-to-end ECG automatic diagnosis through deep learning is essential due to the time-consuming and labour-intensive characteristics of manual diagnosis of ECG records. This paper presents the design of a hybrid residual recurrent convolutional neural network (HRRCNN) for automatically diagnosing single-lead ECG with multiple labels. Firstly, the convolutional layers and residual networks extract the spatial and short-term temporal features. Subsequently, a hybrid recurrent neural network (RNN) comprised of a double-layer bidirectional gated recurrent unit (GRU) and a bidirectional long short-term memory (LSTM) is employed to capture long and short-term temporal features. Finally, an output layer, comprising an attention layer and a fully connected layer, converts the extracted features into classification results. To evaluate the performance of HRRCNN, a dataset from the Shanghai First People’s Hospital Affiliated to Shanghai JiaoTong University(Shanghai General Hospital, SGH) is utilized. The experimental results demonstrate that HRRCNN achieves an accuracy of 95.08 %, F1-macro of 87.22 %, Sensitivity of 88.51 %, and a positive prediction rate of 86.13 %. Comparative analysis with existing methods underscores the superior performance of HRRCNN, thereby establishing its potential as an out-of-hospital heart health monitoring tool and an auxiliary tool in clinical settings.

Translating Gestures: Utilizing CNN to Enhance ASL Communication and Understanding

This research introduces a system for translating American Sign Language (ASL) utilising Convolutional Neural Networks. (CNNs). The system under consideration has the capability to identify manual gestures executed by persons communicating in American Sign Language (ASL) and subsequently convert them into written language, thereby facilitating uninterrupted communication between the deaf and hearing populations. The CNN model was validated and trained using a set of data made up of 78,300 pictures of movements in the ASL Alphabet written in American Sign Language (ASL). In order to improve the performance of the model, the pre-processing of the data included a number of stages, such as converting the photos to grayscale, normalising the pixel values, and enhancing performance of the model. In order to ease classification, fully connected layers were added after a succession of pooling and convolutional layers in CNN's design. After 15 epochs of training, the model attained a validation accuracy of 99.85%. The findings of this research demonstrate the viability of employing Convolutional Neural Networks (CNNs) in the creation of precise and effective American Sign Language (ASL) translation systems, which can serve as a means of communication for people with auditory impairments.

A Machine Learning Application for Medical Image Analysis Using Deep Convolutional Neural Networks (CNNs) and Transfer Learning Models for Pneumonia Detection

Pneumonia is still a major problem in world health, causing a lot of illness and death, particularly among the most susceptible people. For care and therapy to be successful, a prompt and precise diagnosis is essential. Recent years have seen encouraging outcomes in medical image processing tasks when using deep learning methods, including Convolutional Neural Networks (CNNs). Using deep convolutional neural networks (CNNs) and transfer learning models, we provide a machine learning application for detecting pneumonia in medical photographs.Our suggested system makes use of convolutional neural network (CNN) designs that have already been trained using massive natural picture datasets, including VGG, ResNet, and DenseNet. We hope that by training these algorithms on a collection of chest X-ray pictures, we may modify their characteristics to better diagnose pneumonia. Furthermore, we investigate data augmentation methods to improve the models' generalizability, especially in cases when there is a lack of labelled data.Standard measures including precision, sensitivity, specificity, and area under the curve of the receiver operating characteristic (AUC-ROC) are used to assess the performance of the created models on a heterogeneous dataset of chest X-ray images.

Evaluation of the effects of K-Means Clustered-Based Weight Quantization on a Keras Library Based Convolutional Neural Network for Hand Written Digit Image Recognition

A lot of Convolutional Neural Networks (CNNs) have been implemented using FPGAs for the past years. Subsequently, memory saving features were added to the CNN through weight quantization using K-means clustering. A future goal on an ASIC design, involving CNN and weight quantization working together in one chip, can give way to an automated procedure of memory-saving CNN design. In this paper an evaluation was done on the effect of quantizing the weights of a Keras library-based CNN using K means clustering. Various values of K in K-means clustering were tested to see its effects on the CNN accuracy performance. This paper presents first the design approach of a Keras library based Convolutional Neural Network (CNN) for hand-written digit images. It then presents a hardware model design of K-Means clustering algorithm using VHDL. The performance of CNN for image recognition was then tested for various levels of weight quantization using K-means clustering algorithm. Simulation results showed a compression of weights as high as 60% resulted to less than 1% reduction in CNN’s accuracy. The findings in this paper will serve as guide in determining the relevant values of K i.e. the compression ratio, for future ASIC design on this topic.

Investigate the Application of Particle Swarm Optimization to Fine-Tune the Architecture and Parameters of Deep Convolutional Neural Networks for Enhanced Accuracy in Brain Tumor Detection from Medical Images

This study researches the use of Particle Swarm Optimization (PSO) to adjust the design and boundaries of Deep Convolutional Neural Networks (DCNNs) for improved precision in cerebrum growth recognition from clinical pictures. Different PSO calculation variants are investigated for division and grouping of MR pictures. A clever PSO-based strategy for picture gathering is presented, alongside a half and half K-Means/SBM-PSO approach for MR picture division. The proposed methods are assessed utilizing MR pictures from assorted sources, uncovering the adequacy of PSO in both picture division and advancing the K-Means grouping strategy. A half breed PSO approach is exhibited for characterizing MR pictures as typical or strange in light of the presence of mind growths, and for evaluating MR pictures as per the WHO characterization framework for cerebrum cancers. The exploratory outcomes demonstrate that the proposed techniques lead to more modest intra-bunch distances and bigger between-bunch distances, bringing about superior division results. Eminently, the reconciliation of PSO with K-Means shows improved strength and execution contrasted with individual methodologies. The reviewing discoveries recommend impediments in the ongoing X-ray approach for growth evaluating. Overall, this study highlights the potential of PSO in optimizing DCNN architectures for accurate brain tumor detection from medical images, emphasizing the effectiveness of hybrid PSO-K-Means models for improved segmentation and classification outcomes.

Breast Cancer Detection Using Deep Learning

The most frequently occurring cancer among Indian women is breast cancer. The second major cause of women's death is breast cancer (after lung cancer). 246,660 of women's new cases of invasive breast cancer are expected to be diagnosed in the US during 2016 and 40,450 of women’s death is estimated. The number of breast cancer cases is steadily growing especially with increasing number of ageing population. Breast cancer is a type of cancer that starts in the breast. Cancer starts when cells begin to grow out of control. Breast cancer cells usually form a tumour that can often be seen on an x-ray or felt as a lump. The screening practice using mammography needs to be better and potentially efficient. There is always room for advancement when it comes to medical imaging. Early detection of cancer can reduce the risk of death for cancer patients. Breast cancer can spread when the cancer cells get into the blood or lymph system and are carried to other parts of the body. This overview looks at the results of a proposed system for automatically diagnosing breast cancer using several convolutional neural network (CNN) designs and compares them to those achieved using machine learning (ML) techniques. All buildings were built from a large database. Validation tests were conducted in order to provide quantifiable results using the performance parameters of each approach. Functions like VGG16 and ResNet50 are used in this model to improve the accuracy of the model. The cause of Breast Cancer includes changes and mutations in DNA. Women are seriously threatened by breast cancer with high morbidity and mortality. The breast cancer dataset is publicly available on the UCI Machine Learning Repository and therefore the implementation phase dataset is going to be partitioned as 80% for the training phase and 20% for the testing phase then apply a significant performance in respect of all parameters.

Graph Neural Networks optimized with Gazelle Optimization Algorithm for Urban Plantscape Design Based on Large-scale Street View

The center of human activity and a key point of contact for social interaction between city people and the built environment is the street. With more street-level photos becoming available, urban landscape studies have more chances to examine and evaluate streetscapes closely and from various angles. This manuscript present the graph neural networks (GNN) optimized with gazelle optimization algorithm (GOA) for urban plants cape design based on large-scale street view (UPLS-GNN-GOA). Initially data is taken from ADE20K dataset. Afterward the data is fed to General Robust Subband Adaptive Filtering (GRSAF) based pre-processing process. The pre-processing output is given to Discrete Fractional Fourier Transforms (DFrFT) to extract the street greenery and openness of street canyons. After that, the extracted features are fed to Graph Neural Networks (GNN) to classify the public health, urban microclimate, human perception. The weight parameters of the GNN are optimized using Gazelle Optimization Algorithm (GOA).The UPLS-GNN-GOA method is implemented in Python, and its efficiency is determined using a series of performance evaluation metrics analysis, including accuracy, precision, recall, f1-score, and sensitivity. The proposed UPLS-GNN-GOA method shows the highest accuracy of 98%, precision of 98%, and F1-score of 97% while comparing other existing methods such as urban plantscape design based on convolutional neural network (UPLS-CNN), urban plantscape design based on deep convolutional neural network (UPLS-DCNN) and urban plantscape design based on deep learning and convolutional neural network (UPLS-DL-CNN) respectively.

A Comparative Study of AI Methods on Renewable Energy Prediction for Smart Grids: Case of Turkey

Fossil fuels still have emerged as the predominant energy source for power generation on a global scale. In recent years, Turkey has experienced a notable decrease in the production of coal and natural gas energy, juxtaposed with a significant rise in the production of renewable energy sources. The study employed neural networks, ANNs (artificial neural networks), and LSTM (long short-term memory), as well as CNN (convolutional neural network) and hybrid CNN-LSTM designs, to assess Turkey’s energy potential. Real-time outcomes were produced by integrating these models with meteorological data. The objective was to design strategies for enhancing performance by comparing various models of outcomes. The data collected for Turkey as a whole are based on average values. Machine learning approaches were employed to mitigate the error rate seen in the acquired outcomes. Comparisons were conducted across light gradient boosting machine (LightGBM), gradient boosting regressor (GBR), and random forest regressor (RF) techniques, which represent machine learning models, alongside deep learning models. Based on the findings of the comparative analyses, it was determined that the machine learning model, LightGBM, exhibited the most favorable performance in enhancing the accuracy of predictions. Conversely, the hybrid model, CNN-LSTM, had the greatest rate of inaccuracy. This study will serve as a guide for renewable energy researchers, especially in developing countries such as Turkey that have not switched to a smart grid system.

Comparison of transfer-learning for lightweight pre-trained model on image classification

This paper presents a comparative study of the performance of three convolutional neural network (CNN) architectures - EffcientNet-B0, ResNet-50, and AlexNet - for a given image classification task. The study provides a comprehensive investigation of the training process, hardware configurations, training time, and individual model performance. The investigation also assesses the models suitability for different applications. The findings can help both researchers and practitioners select the most suitable model for their specific needs and applications. The paper provides an analysis of each CNN architecture and discusses their strength and weaknesses. The results demonstrate that EffcientNet-B0 achieves the highest accuracy, but its training performance is not optimal. ResNet-50, on the other hand, exhibits high accuracy with efficient training using transfer learning. Finally, ALEXNET provides a baseline for comparison with traditional CNN designs. The paper also highlights the trade-offs involved in selecting a CNN architecture and highlights their relative advantages and disadvantages. The reader is provided with insights into which CNN architecture is most suitable for specific applications based on their requirements.

Best performance with fewest resources: Unveiling the most resource-efficient Convolutional Neural Network for P300 detection with the aid of Explainable AI

Convolutional Neural Networks (CNNs) have shown remarkable prowess in detecting P300, an Event-Related Potential (ERP) crucial in Brain–Computer Interfaces (BCIs). Researchers persistently seek simple and efficient CNNs for P300 detection, exemplified by models like DeepConvNet, EEGNet, and SepConv1D. Noteworthy progress has been made, manifesting in reducing parameters from millions to hundreds while sustaining state-of-the-art performance. However, achieving further simplification or performance improvement beyond SepConv1D appears challenging due to inherent oversimplification. This study explores landmark CNNs and P300 data with the aid of Explainable AI, proposing a simpler yet superior-performing CNN architecture which incorporates (1) precise separable convolution for feature extraction of P300 data, (2) adaptive activation function tailored for P300 data, and (3) customized large learning rate schedules for training P300 data. Termed the Minimalist CNN for P300 detection (P300MCNN), this novel model is characterized by its requirement of the fewest filters and epochs to date, concurrently achieving best performance in cross-subject P300 detection. P300MCNN not only introduces groundbreaking concepts for CNN architectures in P300 detection but also showcases the importance of Explainable AI in demystifying the “black box” design of CNNs.

Classifying electrocardiograph waveforms using trained deep learning neural network based on wavelet representation

<span lang="EN-US">Due to the rise in cardiac patients, an automated system that can identify different heart disorders has been created to lighten and distribute the duty of physicians. This research uses three different electrocardiograph (ECG) signals as indicators of a person's cardiac problems: Normal sinus rhythm (NSR), arrhythmia (ARR), and congestive heart failure (CHF). The continuous wavelet transform (CWT) provides the mechanism for classifying the 190 individual cases of ECG data into a 2-dimensional time-frequency representation. In this paper, the modified GoogLeNet is used for ECG data classification. Using a transfer learning approach and adjustments to parts of the output layers, ECG classification was conducted and the effectiveness of convolutional neural network (CNN) designs was tested. By comparing the results that the optimized neural network and GoogLeNet both had classification accuracy about of 80% and 100%, respectively. The GoogLeNet provide the best result in term of accuracy and training time.</span>

Detection of Covid-19 based on convolutional neural networks using pre-processed chest X-ray images

The global catastrophe known as COVID-19 has shattered the world’s socioeconomic structure. Effective and affordable diagnosis techniques are crucial for better COVID-19 therapy and the eradication of bogus cases. Due to the daily upsurge in cases, hospitals only have a small supply of COVID-19 test kits. The study describes a deep Convolutional Neural Network (CNN) design for categorizing chest x-ray images in the diagnosis of COVID-19. The lack of a substantial, high-quality chest x-ray picture collection made efficient and exact CNN categorization problematic. The dataset has been pre-processed using an image enhancement strategy to provide an effective training dataset for the proposed CNN model to achieve performance. The proposed model achieves 99.73% of accuracy, 98.95% of specificity, 99.47% of precision, 99.62% of sensitivity, and 98.71% of F1 score. A comparative study between the proposed model and numerous CNN-based COVID-19 detection algorithms is carried out to demonstrate that it outperforms other models. When evaluated on a separate dataset, the suggested model excelled over all other models, generally and explicitly.

CNN-LSTM and transfer learning models for malware classification based on opcodes and API calls

In this paper, we propose a novel model for a malware classification system based on Application Programming Interface (API) calls and opcodes, to improve classification accuracy. This system uses a novel design of combined Convolutional Neural Network and Long Short-Term Memory. We extract opcode sequences and API Calls from Windows malware samples for classification. We transform these features into N-grams (N = 2, 3, and 10)-gram sequences. Our experiments on a dataset of 9,749,57 samples produce high accuracy of 99.91% using the 8-gram sequences. Our method significantly improves the malware classification performance when using a wide range of recent deep learning architectures, leading to state-of-the-art performance. In particular, we experiment with ConvNeXt-T, ConvNeXt-S, RegNetY-4GF, RegNetY-8GF, RegNetY-12GF, EfficientNetV2, Sequencer2D-L, Swin-T, ViT-G/14, ViT-Ti, ViT-S, VIT-B, VIT-L, and MaxViT-B. Among these architectures, Swin-T and Sequencer2D-L architectures achieved high accuracies of 99.82% and 99.70%, respectively, comparable to our CNN-LSTM architecture although not surpassing it.

HistoEM: A Pathologist-Guided and Explainable Workflow Using Histogram Embedding for Gland Classification

Pathologists have, over several decades, developed criteria for diagnosing and grading prostate cancer. However, this knowledge has not, so far, been included in the design of convolutional neural networks (CNN) for prostate cancer detection and grading. Further, it is not known whether the features learned by machine-learning algorithms coincide with diagnostic features used by pathologists. We propose a framework that enforces algorithms to learn the cellular and subcellular differences between benign and cancerous prostate glands in digital slides from hematoxylin and eosin–stained tissue sections. After accurate gland segmentation and exclusion of the stroma, the central component of the pipeline, named HistoEM, utilizes a histogram embedding of features from the latent space of the CNN encoder. Each gland is represented by 128 feature-wise histograms that provide the input into a second network for benign vs cancer classification of the whole gland. Cancer glands are further processed by a U-Net structured network to separate low-grade from high-grade cancer. Our model demonstrates similar performance compared with other state-of-the-art prostate cancer grading models with gland-level resolution. To understand the features learned by HistoEM, we first rank features based on the distance between benign and cancer histograms and visualize the tissue origins of the 2 most important features. A heatmap of pixel activation by each feature is generated using Grad-CAM and overlaid on nuclear segmentation outlines. We conclude that HistoEM, similar to pathologists, uses nuclear features for the detection of prostate cancer. Altogether, this novel approach can be broadly deployed to visualize computer-learned features in histopathology images.

Brain tumor prediction and analysis based on simple CNN

The field of brain tumor prediction aims to detect potentially malignant tumors before overt symptoms appear. Early tumor detection can help tailor treatment pathways. To this end, this paper introduces a convolutional neural network (CNN) to construct a simple CNN analysis model. Subtle architectural changes to typical CNN designs are aimed at navigating and solving complex problems unique to medical imaging. Specifically, first, a preprocessing stage converts raw Magnetic Resonance Imaging (MRI) scans into analyzable data. This enables the model to delve into complex patterns in each image. Second, build a simple CNN model and evaluate it. Additionally, comparisons are made with established benchmarks set by traditional CNN architectures. The study further utilizes many available annotated datasets to develop a data augmentation paradigm that improves model resilience and generalizability. At the same time, the study deploys rigorous protocols to address and contain the ubiquitous overfitting threat, thereby enhancing the general applicability of the solution across a variety of healthcare settings. Empirical evaluations shed light on the significant superiority of the proposed simple CNN, especially in specific imaging environments and data structures. At the heart of this research is not only to expand technological horizons but also to reimagine clinical diagnostics to improve the standard of patient care.

CVTStego-Net: A convolutional vision transformer architecture for spatial image steganalysis

The principal investigations in image steganalysis in the spatial domain have concentrated on convolutional neural network (CNN) designs. However, existing CNNs increase the local receptive field of steganographic noise without considering global steganographic noise. This study introduces CVTStego-Net, a convolutional vision transformer for spatial domain image steganalysis that merges the strengths of convolutions and the advantages of attention mechanisms to capture both local and global dependencies. CVTStego-Net is composed of three stages: preprocessing stage, noise extraction, and analysis stage, and classification stage. The preprocessing stage involves a bifurcation with trainable and untrainable 30 SRM (Spatial Rich Models) filters to enhance steganographic noise. The noise extraction and analysis stage combines the SE-Block (Squeeze-and-Excitation) with residual operations to increase the sensitivity to steganographic noise and suppressing the influence of redundant information, and the classification stage combines SE-Block with a convolutional vision transformer to connect the local and global spatial relationships of the steganographic noise. This work enhanced the classification accuracies for steganographic algorithms compared to YEDROUDJ-Net, SR-Net, ZHU-Net, GBRAS-Net, and SNMC-Net. Specifically, the accuracy of CVTStego-Net for WOW at 0.2 bpp was 86.58%, and 0.4 bpp was 93.80%. Moreover, for S-UNIWARD at 0.2 and 0.4 bpp, the accuracies were 80.70% and 90.45%, respectively. For MiPOD at 0.2 and 0.4 bpp, the accuracies were 74.70% and 81.48%, respectively. For HILL at 0.2 and 0.4 bpp, the accuracies were 76.70% and 85.80%, respectively, and for HUGO at 0.2 and 0.4 bpp, the accuracies were 78.20% and 86.98%, respectively, using test data from the BOSSbase 1.01. The results demonstrate that convolutional vision transformers can classify steganographic images in the spatial domain.

PsoResNet: An improved PSO-based residual network search algorithm

Neural Architecture Search (NAS) methods are widely employed to address the time-consuming and costly challenges associated with manual operation and design of deep convolutional neural networks (DCNNs). Nonetheless, prevailing methods still encounter several pressing obstacles, including limited network architecture design, excessively lengthy search periods, and insufficient utilization of the search space. In light of these concerns, this study proposes an optimization strategy for residual networks that leverages an enhanced Particle swarm optimization algorithm. Primarily, low-complexity residual architecture block is employed as the foundational unit for architecture exploration, facilitating a more diverse investigation into network architectures while minimizing parameters. Additionally, we employ a depth initialization strategy to confine the search space within a reasonable range, thereby mitigating unnecessary particle exploration. Lastly, we present a novel approach for computing particle differences and updating velocity mechanisms to enhance the exploration of updated trajectories. This method significantly contributes to the improved utilization of the search space and the augmentation of particle diversity. Moreover, we constructed a crime-dataset comprising 13 classes to assess the effectiveness of the proposed algorithm. Experimental results demonstrate that our algorithm can design lightweight networks with superior classification performance on both benchmark datasets and the crime-dataset.

Detection of COVID-19 using edge devices by a light-weight convolutional neural network from chest X-ray images.

Deep learning is a highly significant technology in clinical treatment and diagnostics nowadays. Convolutional Neural Network (CNN) is a new idea in deep learning that is being used in the area of computer vision. The COVID-19 detection is the subject of our medical study. Researchers attempted to increase the detection accuracy but at the cost of high model complexity. In this paper, we desire to achieve better accuracy with little training space and time so that this model easily deployed in edge devices. In this paper, a new CNN design is proposed that has three stages: pre-processing, which removes the black padding on the side initially; convolution, which employs filter banks; and feature extraction, which makes use of deep convolutional layers with skip connections. In order to train the model, chest X-ray images are partitioned into three sets: learning(0.7), validation(0.1), and testing(0.2). The models are then evaluated using the test and training data. The LMNet, CoroNet, CVDNet, and Deep GRU-CNN models are the other four models used in the same experiment. The propose model achieved 99.47% & 98.91% accuracy on training and testing respectively. Additionally, it achieved 97.54%, 98.19%, 99.49%, and 97.86% scores for precision, recall, specificity, and f1-score respectively. The proposed model obtained nearly equivalent accuracy and other similar metrics when compared with other models but greatly reduced the model complexity. Moreover, it is found that proposed model is less prone to over fitting as compared to other models.