Popular Convolutional Neural Network Architectures Research Articles

Bird Species Detection Using Deep Learning

Bird species classification plays a crucial role in various ecological and conservation endeavors. Leveraging deep learning techniques has shown promise in automating this task, offering the potential for efficient and accurate species identification. This paper presents a comprehensive study on the application of the YOLOv5 architecture, a state-of-the-art object detection framework, for bird species classification from images. We explore the effectiveness of YOLOv5 in comparison to traditional convolutional neural networks (CNNs) for bird species identification. Our study involves extensive experimentation with different configurations of YOLOv5, including variations in model size, training strategies, and dataset preprocessing techniques. We evaluate the performance of the YOLOv5-based classification framework on diverse datasets, ranging from publicly available bird image datasets to custom datasets collected through field observations. Additionally, we compare the classification accuracy of YOLOv5 with other popular CNN architectures, such as ResNet and EfficientNet, to assess its efficacy in bird species recognition tasks. Furthermore, we investigate the transferability of pre-trained YOLOv5 models to different bird species datasets and examine the robustness of the models to variations in image quality, background clutter, and occlusions.Our experimental results demonstrate that YOLOv5 offers competitive performance in bird species classification tasks, achieving high accuracy and efficiency. We discuss the strengths and limitations of using YOLOv5 for this application and provide insights into potential avenues for future research.

Transfer learning with CNNs for efficient prostate cancer and BPH detection in transrectal ultrasound images

Early detection of prostate cancer (PCa) and benign prostatic hyperplasia (BPH) is crucial for maintaining the health and well-being of aging male populations. This study aims to evaluate the performance of transfer learning with convolutional neural networks (CNNs) for efficient classification of PCa and BPH in transrectal ultrasound (TRUS) images. A retrospective experimental design was employed in this study, with 1380 TRUS images for PCa and 1530 for BPH. Seven state-of-the-art deep learning (DL) methods were employed as classifiers with transfer learning applied to popular CNN architectures. Performance indices, including sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV), Kappa value, and Hindex (Youden’s index), were used to assess the feasibility and efficacy of the CNN methods. The CNN methods with transfer learning demonstrated a high classification performance for TRUS images, with all accuracy, specificity, sensitivity, PPV, NPV, Kappa, and Hindex values surpassing 0.9400. The optimal accuracy, sensitivity, and specificity reached 0.9987, 0.9980, and 0.9980, respectively, as evaluated using twofold cross-validation. The investigated CNN methods with transfer learning showcased their efficiency and ability for the classification of PCa and BPH in TRUS images. Notably, the EfficientNetV2 with transfer learning displayed a high degree of effectiveness in distinguishing between PCa and BPH, making it a promising tool for future diagnostic applications.

Hybrid morphological-convolutional neural networks for computer-aided diagnosis.

Training deep Convolutional Neural Networks (CNNs) presents challenges in terms of memory requirements and computational resources, often resulting in issues such as model overfitting and lack of generalization. These challenges can only be mitigated by using an excessive number of training images. However, medical image datasets commonly suffer from data scarcity due to the complexities involved in their acquisition, preparation, and curation. To address this issue, we propose a compact and hybrid machine learning architecture based on the Morphological and Convolutional Neural Network (MCNN), followed by a Random Forest classifier. Unlike deep CNN architectures, the MCNN was specifically designed to achieve effective performance with medical image datasets limited to a few hundred samples. It incorporates various morphological operations into a single layer and uses independent neural networks to extract information from each signal channel. The final classification is obtained by utilizing a Random Forest classifier on the outputs of the last neural network layer. We compare the classification performance of our proposed method with three popular deep CNN architectures (ResNet-18, ShuffleNet-V2, and MobileNet-V2) using two training approaches: full training and transfer learning. The evaluation was conducted on two distinct medical image datasets: the ISIC dataset for melanoma classification and the ORIGA dataset for glaucoma classification. Results demonstrate that the MCNN method exhibits reliable performance in melanoma classification, achieving an AUC of 0.94 (95% CI: 0.91 to 0.97), outperforming the popular CNN architectures. For the glaucoma dataset, the MCNN achieved an AUC of 0.65 (95% CI: 0.53 to 0.74), which was similar to the performance of the popular CNN architectures. This study contributes to the understanding of mathematical morphology in shallow neural networks for medical image classification and highlights the potential of hybrid architectures in effectively learning from medical image datasets that are limited by a small number of case samples.

Convolutional Neural Networks: A Survey

Artificial intelligence (AI) has become a cornerstone of modern technology, revolutionizing industries from healthcare to finance. Convolutional neural networks (CNNs) are a subset of AI that have emerged as a powerful tool for various tasks including image recognition, speech recognition, natural language processing (NLP), and even in the field of genomics, where they have been utilized to classify DNA sequences. This paper provides a comprehensive overview of CNNs and their applications in image recognition tasks. It first introduces the fundamentals of CNNs, including the layers of CNNs, convolution operation (Conv_Op), Feat_Maps, activation functions (Activ_Func), and training methods. It then discusses several popular CNN architectures such as LeNet, AlexNet, VGG, ResNet, and InceptionNet, and compares their performance. It also examines when to use CNNs, their advantages and limitations, and provides recommendations for developers and data scientists, including preprocessing the data, choosing appropriate hyperparameters (Hyper_Param), and evaluating model performance. It further explores the existing platforms and libraries for CNNs such as TensorFlow, Keras, PyTorch, Caffe, and MXNet, and compares their features and functionalities. Moreover, it estimates the cost of using CNNs and discusses potential cost-saving strategies. Finally, it reviews recent developments in CNNs, including attention mechanisms, capsule networks, transfer learning, adversarial training, quantization and compression, and enhancing the reliability and efficiency of CNNs through formal methods. The paper is concluded by summarizing the key takeaways and discussing the future directions of CNN research and development.

Image-Based Arabian Camel Breed Classification Using Transfer Learning on CNNs

Image-based Arabian camel breed classification is an important task for various practical applications, such as breeding management, genetic improvement, conservation, and traceability. However, it is a challenging task due to the lack of standardized criteria and methods, the high similarity among breeds, and the limited availability of data and resources. In this paper, we propose an approach to tackle this challenge by using convolutional neural networks (CNNs) and transfer learning to classify images of six different Arabian camel breeds: Waddeh, Majaheem, Homor, Sofor, Shaele, and Shageh. To achieve this, we created, preprocessed, and annotated a novel dataset of 1073 camel images. We then pre-trained CNNs as feature extractors and fine-tuned them on our new dataset. We evaluated several popular CNN architectures with diverse characteristics such as InceptionV3, NASNetLarge, PNASNet-5-Large, MobileNetV3-Large, and EfficientNetV2 (small, medium, and large variants), and we found that NASNetLarge achieves the best test accuracy of 85.80% on our proposed dataset. Finally, we integrated the best-performing CNN architecture, NASNetLarge, into a mobile application for further validation and actual use in a real-world scenarios.

Archeological Sites Classification Through Partial Imaging and Convolutional Neural Networks

In this paper, a novel approach for classifying archeological sites using publicly available images through the use of Convolutional Neural Networks (CNNs) is presented. To surmount the problem of having a limited amount of data to use in training and testing the CNNs, our approach employs the technique of fine tuning. We conducted an experiment with four popular CNN architectures: VGG-16, VGG-19, ResNet50, and InceptionV3. The results show that our models achieved an impressive accuracy of up to 98% using the VGG-16 and InceptionV3 models and up to 97% using the ResNet50 model, while the VGG-19 model produced results with an accuracy of 95%. The results of this study demonstrate the effectiveness of our proposed approach in classifying archeological sites using publicly available images and highlight the potential of deep learning techniques for archeological site classification.

Investigating the Impact of Two Major Programming Environments on the Accuracy of Deep Learning-Based Glioma Detection from MRI Images

Brain tumors have been the subject of research for many years. Brain tumors are typically classified into two main groups: benign and malignant tumors. The most common tumor type among malignant brain tumors is known as glioma. In the diagnosis of glioma, different imaging technologies could be used. Among these techniques, MRI is the most preferred imaging technology due to its high-resolution image data. However, the detection of gliomas from a huge set of MRI data could be challenging for the practitioners. In order to solve this concern, many Deep Learning (DL) models based on Convolutional Neural Networks (CNNs) have been proposed to be used in detecting glioma. However, understanding which CNN architecture would work efficiently under various conditions including development environment or programming aspects as well as performance analysis has not been studied so far. In this research work, therefore, the purpose is to investigate the impact of two major programming environments (namely, MATLAB and Python) on the accuracy of CNN-based glioma detection from Magnetic Resonance Imaging (MRI) images. To this end, experiments on the Brain Tumor Segmentation (BraTS) dataset (2016 and 2017) consisting of multiparametric magnetic MRI images are performed by implementing two popular CNN architectures, the three-dimensional (3D) U-Net and the V-Net in the programming environments. From the results, it is concluded that the use of Python with Google Colaboratory (Colab) might be highly useful in the implementation of CNN-based models for glioma detection. Moreover, the 3D U-Net model is found to perform better, attaining a high accuracy on the dataset. The authors believe that the results achieved from this study would provide useful information to the research community in their appropriate implementation of DL approaches for brain tumor detection.

Road crack avoidance: a convolutional neural network-based smart transportation system for intelligent vehicles

Prediction using computer vision is getting prevalent nowadays because of satisfying results. The vision of Internet of Vehicles (IoV) expedites Vehicle to everything (V2X) communications by implementing heterogeneous global networks. Road crack is one of the major factors that causes road mishaps and damage to vehicles. To ensure smooth and safe driving, avoiding road crack in transportation planning and navigation is significant. To address this issue, we proposed a novel convolutional neural network (CNN)-based smart transportation system. We showed how to quantify the severity of the cracks. We proposed a post-processing algorithm to provide option to the driver to select the safest road toward the destination. The communication system for the proposed smart transportation system has also been introduced. The performance comparison of a few popular CNN architectures has been investigated. Simulation results showed that Resnet50 algorithm provides significantly high accuracy compared with SqueezeNet and InceptionV3 algorithm in order to detect road cracks for the proposed transportation system. We demonstrated high accuracy of measuring the crack severity via numerical analysis. The integration of the proposed system in next generation smart vehicles can ensure accurate detection of road cracks earlier enough providing the option to select alternate safe route toward a destination as advanced driver assistance service. Moreover, the proposed system can also play a key role in order to reduce road mishaps notably by warning the driver about the updated road surface conditions.

Transfer Learning Approach for Human Activity Recognition Based on Continuous Wavelet Transform

Over the last few years, human activity recognition (HAR) has drawn increasing interest from the scientific community. This attention is mainly attributable to the proliferation of wearable sensors and the expanding role of HAR in such fields as healthcare, sports, and human activity monitoring. Convolutional neural networks (CNN) are becoming a popular approach for addressing HAR problems. However, this method requires extensive training datasets to perform adequately on new data. This paper proposes a novel deep learning model pre-trained on scalograms generated using the continuous wavelet transform (CWT). Nine popular CNN architectures and different CWT configurations were considered to select the best performing combination, resulting in the training and evaluation of more than 300 deep learning models. On the source KU-HAR dataset, the selected model achieved classification accuracy and an F1 score of 97.48% and 97.52%, respectively, which outperformed contemporary state-of-the-art works where this dataset was employed. On the target UCI-HAPT dataset, the proposed model resulted in a maximum accuracy and F1-score increase of 0.21% and 0.33%, respectively, on the whole UCI-HAPT dataset and of 2.82% and 2.89%, respectively, on the UCI-HAPT subset. It was concluded that the usage of the proposed model, particularly with frozen layers, results in improved performance, faster training, and smoother gradient descent on small HAR datasets. However, the use of the pre-trained model on sufficiently large datasets may lead to negative transfer and accuracy degradation.

An in-depth analysis of Convolutional Neural Network architectures with transfer learning for skin disease diagnosis

Low contrasts and visual similarity between different skin conditions make skin disease recognition a challenging task. Current techniques to detect and diagnose skin disease accurately require high-level professional expertise. Artificial intelligence paves the way for developing computer vision-based applications in medical imaging, like recognizing dermatological conditions. This research proposed an efficient solution for skin disease recognition by implementing Convolutional Neural Network (CNN) architectures. Computer vision-based applications using CNN architectures, MobileNet and Xception, are used to construct an expert system that can accurately and efficiently recognize different classes of skin diseases accurately and efficiently. The proposed CNN architectures used a transfer learning method in which models are pre-trained on the Imagenet dataset to discover more features. We also evaluated the performance of our proposed approach with some of the most popular CNN architectures: ResNet50, InceptionV3, Inception-ResNet, and DenseNet, thus establishing a comparison to set up a benchmark that will ratify the essence of transfer learning and augmentation. This study uses data from two separate data sources to collect five different types of skin disorders. Different performance evaluation indicators, including accuracy, precision, recall, and F1-score, are calculated to verify the success of our technique. The experimental results revealed the effectiveness of our proposed approach, where MobileNet achieved a classification accuracy of 96.00%, and the Xception model reached 97.00% classification accuracy with transfer learning and augmentation. Moreover, we proposed and implemented a web-based architecture for the real-time recognition of diseases.

Understanding of Convolutional Neural Network (CNN): A Review

The application of deep learning technology has increased rapidly in recent years. Technologies in deep learning increasingly emulate natural human abilities, such as knowledge learning, problem-solving, and decision-making. In general, deep learning can carry out self-training without repetitive programming by humans. Convolutional neural networks (CNNs) are deep learning algorithms commonly used in wide applications. CNN is often used for image classification, segmentation, object detection, video processing, natural language processing, and speech recognition. CNN has four layers: convolution layer, pooling layer, fully connected layer, and non-linear layer. The convolutional layer uses kernel filters to calculate the convolution of the input image by extracting the fundamental features. The pooling layer combines two successive convolutional layers. The third layer is the fully connected layer, commonly called the convolutional output layer. The activation function defines the output of a neural network, such as 'yes' or 'no'. The most common and popular CNN activation functions are Sigmoid, Tanh, ReLU, Leaky ReLU, Noisy ReLU, and Parametric Linear Units. The organization and function of the visual cortex greatly influence CNN architecture because it is designed to resemble the neuronal connections in the human brain. Some of the popular CNN architectures are LeNet, AlexNet and VGGNet.

A Deep Transfer Learning-Based Comparative Study for Detection of Malaria Disease

Malaria is a disease caused by a parasite. The parasite is transmitted to humans through the bite of infected mosquitoes. Thousands of people die every year due to malaria. When this disease is diagnosed early, it can be fully treated with medication. Diagnosis of malaria can be made according to the presence of parasites in the blood taken from the patient. In this study, malaria detection and diagnosis study were performed using The Malaria dataset containing a total of 27,558 cell images with samples of equally parasitized and uninfected cells from thin blood smear slide images of segmented cells. It is possible to detect malaria from microscopic blood smear images via modern deep learning techniques. In this study, 5 of the popular convolutional neural network architectures for malaria detection from cell images were retrained to find the best combination of architecture and learning algorithm. AlexNet, GoogLeNet, ResNet-50, MobileNet-v2, VGG-16 architectures from pre-trained networks were used, their hyperparameters were adjusted and their performances were compared. In this study, a maximum 96.53% accuracy rate was achieved with MobileNet-v2 architecture using the adam learning algorithm

Metamorphic Testing of Image Classification and Consistency Analysis Using Clustering

Testing deep learning systems requires expensive labeled data. In recent years, researchers began to leverage metamorphic testing to address this issue. However, metamorphic relations on image data remain poorly understood. To gain a deeper understanding of these metamorphic relations, we survey common image operations modeling covariate shift, manually classify and categorize the underlying metamorphic relations, and conduct experiments to validate our classifications. In our experiments, we train three popular convolutional neural network architectures on an image classification task. Next, we apply metamorphic operations on input test images and measure the change in classification accuracy and cross-entropy loss. A hierarchical clustering algorithm cluster these results and plots a dendrogram. We compare the groups from manual classification and the clusters from the algorithm to provide key insights. We find that Affine and Noise relations are consistent. Furthermore, we recommend metamorphic relationships to save time and better test deep learning systems in the future.

Building Efficient CNN Architectures for Histopathology Images Analysis: A Case-Study in Tumor-Infiltrating Lymphocytes Classification.

BackgroundDeep learning methods have demonstrated remarkable performance in pathology image analysis, but they are computationally very demanding. The aim of our study is to reduce their computational cost to enable their use with large tissue image datasets.MethodsWe propose a method called Network Auto-Reduction (NAR) that simplifies a Convolutional Neural Network (CNN) by reducing the network to minimize the computational cost of doing a prediction. NAR performs a compound scaling in which the width, depth, and resolution dimensions of the network are reduced together to maintain a balance among them in the resulting simplified network. We compare our method with a state-of-the-art solution called ResRep. The evaluation is carried out with popular CNN architectures and a real-world application that identifies distributions of tumor-infiltrating lymphocytes in tissue images.ResultsThe experimental results show that both ResRep and NAR are able to generate simplified, more efficient versions of ResNet50 V2. The simplified versions by ResRep and NAR require 1.32× and 3.26× fewer floating-point operations (FLOPs), respectively, than the original network without a loss in classification power as measured by the Area under the Curve (AUC) metric. When applied to a deeper and more computationally expensive network, Inception V4, NAR is able to generate a version that requires 4× lower than the original version with the same AUC performance.ConclusionsNAR is able to achieve substantial reductions in the execution cost of two popular CNN architectures, while resulting in small or no loss in model accuracy. Such cost savings can significantly improve the use of deep learning methods in digital pathology. They can enable studies with larger tissue image datasets and facilitate the use of less expensive and more accessible graphics processing units (GPUs), thus reducing the computing costs of a study.

Subdivision-based Mesh Convolution Networks

Convolutionalneural networks (CNNs) have made great breakthroughs in two-dimensional (2D) computer vision. However, their irregular structure makes it hard to harness the potential of CNNs directly on meshes. A subdivision surface provides a hierarchical multi-resolution structure in which each face in a closed 2-manifold triangle mesh is exactly adjacent to three faces. Motivated by these two observations, this article presents SubdivNet , an innovative and versatile CNN framework for three-dimensional (3D) triangle meshes with Loop subdivision sequence connectivity. Making an analogy between mesh faces and pixels in a 2D image allows us to present a mesh convolution operator to aggregate local features from nearby faces. By exploiting face neighborhoods, this convolution can support standard 2D convolutional network concepts, e.g., variable kernel size, stride, and dilation. Based on the multi-resolution hierarchy, we make use of pooling layers that uniformly merge four faces into one and an upsampling method that splits one face into four. Thereby, many popular 2D CNN architectures can be easily adapted to process 3D meshes. Meshes with arbitrary connectivity can be remeshed to have Loop subdivision sequence connectivity via self-parameterization, making SubdivNet a general approach. Extensive evaluation and various applications demonstrate SubdivNet’s effectiveness and efficiency.

Coffee Maturity Classification Using Convolutional Neural Networks and Transfer Learning

This work presents a framework for coffee maturity classification from multispectral image data based on Convolutional Neural Networks (CNNs). The system leverages the use of multispectral image acquisition systems that generate large amounts of data, by taking advantage of the ability of CNNs to extract meaningful patterns from very high-dimensional data. We validated the use of five different popular CNN architectures on the classification of cherry coffee fruits according to their ripening stage. The different models were trained on a training dataset balanced in different ways, which resulted in a top accuracy higher than 98% when applied to the classification of 600 coffee fruits in 5 different stages of ripening. This work has the potential of providing the farmer with a high-quality, optimized, accurate and viable method for classifying coffee fruits. In order to foster future research in this area, the data used in this work, which was acquired with a custom-developed multispectral image acquisition system, have been released.

NemaNet: A convolutional neural network model for identification of soybean nematodes

Phytoparasitic nematodes (or phytonematodes) are causing severe damage to crops and generating large-scale economic losses worldwide. In soybean crops, annual losses are estimated at 10.6% of the world production. Besides, the identification of these species through microscopic analysis by an expert with taxonomic knowledge is often laborious, time-consuming, and susceptible to failure. From this perspective, robust and automatic approaches are necessary for identifying phytonematodes that are capable of providing correct diagnoses for the classification of species and subsidizing of all control and prevention measures. This work presents a new public data set called NemaDataset containing 3063 microscopic images from five nematode species with the most significant damage relevance for the soybean crop. Additionally, we propose a new Convolutional Neural Network (CNN) model defined as NemaNet and present a comparative assessment with thirteen popular models of CNNs, all of them representing state-of-the art classification and recognition. The general average was calculated for each model, on a from-scratch training; the NemaNet model reached 96.76% accuracy, while the best evaluation fold reached 98.04%. When training with transfer learning was performed, the average accuracy reached 98.82%. The best evaluation fold reached 99.35%, and overall accuracy improvements of over 6.83% and 4.1%, for from-scratch and transfer learning training, respectively, compared to other popular models were achieved. • Detect phytonematodes in soybean crops in Brazil using convolutional neural network. • A new dataset with five main species of phytonematodes from Brazil's soybean culture. • Evaluation of thirteen most popular CNN's architectures and new model (NemaNet). • Results describe NemaNet reaches best results in identifying nematodes overall.

Automatic identification of suspicious bone metastatic lesions in bone scintigraphy using convolutional neural network

BackgroundWe aimed to construct an artificial intelligence (AI) guided identification of suspicious bone metastatic lesions from the whole-body bone scintigraphy (WBS) images by convolutional neural networks (CNNs).MethodsWe retrospectively collected the 99mTc-MDP WBS images with confirmed bone lesions from 3352 patients with malignancy. 14,972 bone lesions were delineated manually by physicians and annotated as benign and malignant. The lesion-based differentiating performance of the proposed network was evaluated by fivefold cross validation, and compared with the other three popular CNN architectures for medical imaging. The average sensitivity, specificity, accuracy and the area under receiver operating characteristic curve (AUC) were calculated. To delve the outcomes of this study, we conducted subgroup analyses, including lesion burden number and tumor type for the classifying ability of the CNN.ResultsIn the fivefold cross validation, our proposed network reached the best average accuracy (81.23%) in identifying suspicious bone lesions compared with InceptionV3 (80.61%), VGG16 (81.13%) and DenseNet169 (76.71%). Additionally, the CNN model's lesion-based average sensitivity and specificity were 81.30% and 81.14%, respectively. Based on the lesion burden numbers of each image, the area under the receiver operating characteristic curve (AUC) was 0.847 in the few group (lesion number n ≤ 3), 0.838 in the medium group (n = 4–6), and 0.862 in the extensive group (n > 6). For the three major primary tumor types, the CNN-based lesion identifying AUC value was 0.870 for lung cancer, 0.900 for prostate cancer, and 0.899 for breast cancer.ConclusionThe CNN model suggests potential in identifying suspicious benign and malignant bone lesions from whole-body bone scintigraphic images.

Arrhythmic Heartbeat Classification Using 2D Convolutional Neural Networks

BackgroundElectrocardiogram (ECG) is a method of recording the electrical activity of the heart and it provides a diagnostic means for heart-related diseases. Arrhythmia is any irregularity of the heartbeat that causes an abnormality in the heart rhythm. Early detection of arrhythmia has great importance to prevent many diseases. Manual analysis of ECG recordings is not practical for quickly identifying arrhythmias that may cause sudden deaths. Hence, many studies have been presented to develop computer-aided-diagnosis (CAD) systems to automatically identify arrhythmias. MethodsThis paper proposes a novel deep learning approach to identify arrhythmias in ECG signals. The proposed approach identifies arrhythmia classes using Convolutional Neural Network (CNN) trained by two-dimensional (2D) ECG beat images. Firstly, ECG signals, which consist of 5 different arrhythmias, are segmented into heartbeats which are transformed into 2D grayscale images. Afterward, the images are used as input for training a new CNN architecture to classify heartbeats. ResultsThe experimental results show that the classification performance of the proposed approach reaches an overall accuracy of 99.7%, sensitivity of 99.7%, and specificity of 99.22% in the classification of five different ECG arrhythmias. Further, the proposed CNN architecture is compared to other popular CNN architectures such as LeNet and ResNet-50 to evaluate the performance of the study. ConclusionsTest results demonstrate that the deep network trained by ECG images provides outstanding classification performance of arrhythmic ECG signals and outperforms similar network architectures. Moreover, the proposed method has lower computational costs compared to existing methods and is more suitable for mobile device-based diagnosis systems as it does not involve any complex preprocessing process. Hence, the proposed approach provides a simple and robust automatic cardiac arrhythmia detection scheme for the classification of ECG arrhythmias.

CovidXrayNet: Optimizing data augmentation and CNN hyperparameters for improved COVID-19 detection from CXR

To mitigate the spread of the current coronavirus disease 2019 (COVID-19) pandemic, it is crucial to have an effective screening of infected patients to be isolated and treated. Chest X-Ray (CXR) radiological imaging coupled with Artificial Intelligence (AI) applications, in particular Convolutional Neural Network (CNN), can speed the COVID-19 diagnostic process. In this paper, we optimize the data augmentation and the CNN hyperparameters for detecting COVID-19 from CXRs in terms of validation accuracy. This optimization increases the accuracy of the popular CNN architectures such as the Visual Geometry Group network (VGG-19) and the Residual Neural Network (ResNet-50), by 11.93% and 4.97%, respectively. We then proposed CovidXrayNet model that is based on EfficientNet-B0 and our optimization results. We evaluated CovidXrayNet on two datasets, including our generated balanced COVIDcxr dataset (960 CXRs) and the benchmark COVIDx dataset (15,496 CXRs). With only 30 epochs of training, CovidXrayNet achieves state-of-the-art accuracy of 95.82% on the COVIDx dataset in the three-class classification task (COVID-19, normal or pneumonia). The CovidXRayNet model, the COVIDcxr dataset, and several optimization experiments are publicly available at https://github.com/MaramMonshi/CovidXrayNet.