Deep Learning-based Feature Research Articles

Intelligent Matching Method for Heterogeneous Remote Sensing Images Based on Style Transfer

Intelligent matching of heterogeneous remote sensing images is a common basic problem in the field of intelligent remote sensing image processing. Aiming at the difficulty of matching satellite-aerial remote sensing images, this paper proposes an intelligent matching method for heterogeneous remote sensing images based on style transfer. First, based on the idea of image style transfer of a generative adversarial networks, this method improves the conversion effect of the model on heterogeneous images by constructing a new generative network loss function and converts satellite images into aerial images. Then, the advanced deep learning-based matching Algorithms D2-Net and LoFTR are used to achieve matching between the generated aerial image and the original aerial image. Finally, this transformation relationship is mapped to the corresponding satellite-aerial image pair to obtain the final matching result. The image style transfer experiments and the matching experiments we carry out under different test datasets show that the smooth cycle-consistent generative adversarial networks proposed in this paper can effectively reduce the complexity of the algorithm and improve the quality of image generation. In addition, combining it with deep learning-based feature matching methods can effectively improve the accuracy and robustness of the matching algorithm. Our code and data can be found at: <uri>https://gitee.com/AZQZ/intelligent-matching</uri>.

A Novel Hybrid Approach Based on Deep CNN to Detect Glaucoma Using Fundus Imaging

Glaucoma is one of the eye diseases stimulated by the fluid pressure that increases in the eyes, damaging the optic nerves and causing partial or complete vision loss. As Glaucoma appears in later stages and it is a slow disease, detailed screening and detection of the retinal images is required to avoid vision forfeiture. This study aims to detect glaucoma at early stages with the help of deep learning-based feature extraction. Retinal fundus images are utilized for the training and testing of our proposed model. In the first step, images are pre-processed, before the region of interest (ROI) is extracted employing segmentation. Then, features of the optic disc (OD) are extracted from the images containing optic cup (OC) utilizing the hybrid features descriptors, i.e., convolutional neural network (CNN), local binary patterns (LBP), histogram of oriented gradients (HOG), and speeded up robust features (SURF). Moreover, low-level features are extracted using HOG, whereas texture features are extracted using the LBP and SURF descriptors. Furthermore, high-level features are computed using CNN. Additionally, we have employed a feature selection and ranking technique, i.e., the MR-MR method, to select the most representative features. In the end, multi-class classifiers, i.e., support vector machine (SVM), random forest (RF), and K-nearest neighbor (KNN), are employed for the classification of fundus images as healthy or diseased. To assess the performance of the proposed system, various experiments have been performed using combinations of the aforementioned algorithms that show the proposed model based on the RF algorithm with HOG, CNN, LBP, and SURF feature descriptors, providing ≤99% accuracy on benchmark datasets and 98.8% on k-fold cross-validation for the early detection of glaucoma.

Machine Learning-Based Lie Detector Applied to a Novel Annotated Game Dataset

Lie detection is considered a concern for everyone in their day-to-day life, given its impact on human interactions. Thus, people normally pay attention to both what their interlocutors are saying and to their visual appearance, including the face, to find any signs that indicate whether or not the person is telling the truth. While automatic lie detection may help us to understand these lying characteristics, current systems are still fairly limited, partly due to lack of adequate datasets to evaluate their performance in realistic scenarios. In this work, we collect an annotated dataset of facial images, comprising both 2D and 3D information of several participants during a card game that encourages players to lie. Using our collected dataset, we evaluate several types of machine learning-based lie detectors in terms of their generalization, in person-specific and cross-application experiments. We first extract both handcrafted and deep learning-based features as relevant visual inputs, then pass them into multiple types of classifier to predict respective lie/non-lie labels. Subsequently, we use several metrics to judge the models’ accuracy based on the models predictions and ground truth. In our experiment, we show that models based on deep learning achieve the highest accuracy, reaching up to 57% for the generalization task and 63% when applied to detect the lie to a single participant. We further highlight the limitation of the deep learning-based lie detector when dealing with cross-application lie detection tasks. Finally, this analysis along the proposed datasets would potentially be useful not only from the perspective of computational systems perspective (e.g., improving current automatic lie prediction accuracy), but also for other relevant application fields, such as health practitioners in general medical counselings, education in academic settings or finance in the banking sector, where close inspections and understandings of the actual intentions of individuals can be very important.

Intratumoral analysis of digital breast tomosynthesis for predicting the Ki-67 level in breast cancer: A multi-center radiomics study.

To non-invasively evaluate the Ki-67 level in digital breast tomosynthesis (DBT) images of breast cancer (BC) patients based on subregional radiomics. A total of 266 patients who underwent DBT scans were consecutively enrolled at two centers, between September 2017 and September 2021. The whole tumor region was partitioned into various intratumoral subregions, based on individual- and population-level clustering. Handcrafted radiomics and deep learning-based features were extracted from the subregions and from the whole tumor region, and were selected by least absolute shrinkage and selection operator (LASSO) regression, yielding radiomics signatures (RSs). The area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity were calculated to assess the developed RSs. Each breast tumor region was partitioned into an inner subregion (S1) and a marginal subregion (S2). The RSs derived from S1 always generated higher AUCs compared with those from S2 or from the whole tumor region (W), for the external validation cohort (AUCs, S1 vs. W, handcrafted RSs: 0.583 [95% CI, 0.429-0.727] vs. 0.559 [95% CI, 0.405-0.705], p-value: 0.920; deep RSs: 0.670 [95% CI, 0.516-0.802] vs. 0.551 [95% CI, 0.397-0.698], p-value: 0.776). The fusion RSs, combining handcrafted and deep learning-based features derived from S1, yielded the highest AUCs of 0.820 (95% CI, 0.714-0.900) and 0.792 (95% CI, 0.647-0.897) for the internal and external validation cohorts, respectively. The subregional radiomics approach can accurately predict the Ki-67 level based on DBT data; thus, it may be used as a potential non-invasive tool for preoperative treatment planning in BC.

Proficient Masked Face Recognition Method Using Deep Learning Convolution Neural Network in Covid-19 Pandemic

The COVID-19 pandemic is an incomparable disaster triggering massive fatalities and security glitches. Under the pressure of these black clouds public frequently wear masks as safeguard to their lives. Facial Recognition becomes a challenge because significant portion of human face is hidden behind mask. Primarily researchers focus to derive up with recommendations to tackle this problem through prompt and effective solution in this COVID-19 pandemic. This paper presents a trustworthy method to for the recognition of masked faces on un-occluded and deep learning-based features. The first stage is to capture the non-obstructed face region. Then we extract the most significant features from the attained regions (forehead and eye) through pre-trained deep learning CNN. Bag-of- word paradigm to has been applied to the feature maps to quantize them and to get a minor illustration comparing to the CNN’s fully connected layer. In the end a Multilayer Perceptron has been used for classification. High recognition performance with significant accuracy is seen in experimental results.

MRI-based intratumoral and peritumoral radiomics on prediction of lymph-vascular space invasion in cervical cancer: A multi-center study

PurposeTo explore and externally validate intratumoral and peritumoral radiomics on predicting lymph-vascular space invasion (LVSI) in cervical cancer. MethodsA primary cohort (160 patients) was used to develop radiomics models. A consecutively enrolled internal validation cohort (44 patients) and an external validation cohort (36 patients) were used to test the models. The tumor was partitioned into two intratumoral subregions (S1 and S2) based on patient- and population-level clustering. Handcrafted and deep learning-based features were extracted and selected based on each subregion and the whole tumor, and used to build the multi-regional radiomics signature. Prediction capabilities of various machine learning classifiers were compared. A clinical-radiomics nomogram was constructed integrating the multi-regional radiomics signature and the most important clinical predictor. Receiver operating characteristic (ROC), calibration and decision curves were plotted to assess the radiomics models on the time-independent internal and external validation cohorts. ResultsPredictive performance of S1 and S2 both outperformed the whole tumor. The peritumoral region with 6 mm outside the tumor (Peri-6) showed good predictive capability. The multi-regional radiomics signature integrating S1, S2 and Peri-6 yielded AUCs of 0.841, 0.795, 0.817 and 0.815 in the training, test, internal validation and external validation cohort. The nomogram integrating the multi-regional radiomics signature and degree of differentiation achieved the highest AUCs of 0.859, 0.832, 0.835 and 0.825 in the training, test, internal validation and external validation cohort. ConclusionsThe proposed radiomics models combined intratumoral and peritumoral can effectively predict the LVSI in cervical cancer, and may help to improve clinical decision-making.

Efficient masked face recognition method during the COVID-19 pandemic.

The coronavirus disease (COVID-19) is an unparalleled crisis leading to a huge number of casualties and security problems. In order to reduce the spread of coronavirus, people often wear masks to protect themselves. This makes face recognition a very difficult task since certain parts of the face are hidden. A primary focus of researchers during the ongoing coronavirus pandemic is to come up with suggestions to handle this problem through rapid and efficient solutions. In this paper, we propose a reliable method based on occlusion removal and deep learning-based features in order to address the problem of the masked face recognition process. The first step is to remove the masked face region. Next, we apply three pre-trained deep Convolutional Neural Networks (CNN), namely VGG-16, AlexNet, and ResNet-50, and use them to extract deep features from the obtained regions (mostly eyes and forehead regions). The Bag-of-features paradigm is then applied to the feature maps of the last convolutional layer in order to quantize them and to get a slight representation comparing to the fully connected layer of classical CNN. Finally, Multilayer Perceptron (MLP) is applied for the classification process. Experimental results on Real-World-Masked-Face-Dataset show high recognition performance compared to other state-of-the-art methods.

Fusion of Deep Features With Superpixel Based Local Handcrafted Features for Surface Condition Assessment of Metal Oxide Surge Arrester Using Infrared Thermal Images

This letter presents an innovative approach to sense surface contamination severity of in-service metal oxide surge arrester through combining handcrafted and deep learning-based features. For this purpose, different color regions in the captured infrared thermal (IRT) images are isolated to generate superpixel regions through grouping of informative and meaningful pixels. Thereafter, handcrafted color features are extracted from each superpixel region. Additionally, the IRT images are analyzed through pretrained “ResNet50” deep architecture, and deep features are extracted. Thereafter, handcrafted color features and deep features are combined and fed to a classifier for classification of different surface contamination severity. The findings of this letter indicate that the proposed fusion strategy strongly boosts the classification performance compared to each stand-alone approach. Therefore, proposed framework with feature fusion strategy can be implemented for surface condition assessment of metal oxide surge arrester is service.

Feature Level Combination for Face Recognition Based on Convolutional Neural Networks

Face detection and recognition systems have recently achieved encouraging results using deep learning, especially Convolutional Neural Network (CNN). Face Recognition (FR) systems have many challenges in unconstrained environments that decrease the accuracy; for overcoming these challenges, a deep learning-based features combination has been proposed. The scheme performs feature-level combination by applying two pre-trained GoogLeNet and VggNet-16 models as deep feature extractors. First, faces are detected and aligned using the Multi-Task Convolutional Neural Networks (MTCNN) face detector. The deep features are extracted from a face image using each individually pre-trained CNN. Second, features obtained from GoogLeNet and VggNet-16 models are combined using the serial-feature combination method. Finally, a classification task is performed using a multiclass Support Vector Machine (SVM) classifier. Experiments on the following datasets: VggFace2, LFW, Essex, and ORL, indicate the efficacy of the proposed system as the combination of the two pre-trained CNN models improves performance. The combination strategy, in particular, yields an accuracy of 95.33% to 99.29% on all datasets. The proposed system was compared to existing models in terms of the LFW, and ORL datasets, the findings showed that the proposed system outperformed most current models in terms of accuracy.

A Novel Hybrid Approach Based on Deep CNN Features to Detect Knee Osteoarthritis.

In the recent era, various diseases have severely affected the lifestyle of individuals, especially adults. Among these, bone diseases, including Knee Osteoarthritis (KOA), have a great impact on quality of life. KOA is a knee joint problem mainly produced due to decreased Articular Cartilage between femur and tibia bones, producing severe joint pain, effusion, joint movement constraints and gait anomalies. To address these issues, this study presents a novel KOA detection at early stages using deep learning-based feature extraction and classification. Firstly, the input X-ray images are preprocessed, and then the Region of Interest (ROI) is extracted through segmentation. Secondly, features are extracted from preprocessed X-ray images containing knee joint space width using hybrid feature descriptors such as Convolutional Neural Network (CNN) through Local Binary Patterns (LBP) and CNN using Histogram of oriented gradient (HOG). Low-level features are computed by HOG, while texture features are computed employing the LBP descriptor. Lastly, multi-class classifiers, that is, Support Vector Machine (SVM), Random Forest (RF), and K-Nearest Neighbour (KNN), are used for the classification of KOA according to the Kellgren–Lawrence (KL) system. The Kellgren–Lawrence system consists of Grade I, Grade II, Grade III, and Grade IV. Experimental evaluation is performed on various combinations of the proposed framework. The experimental results show that the HOG features descriptor provides approximately 97% accuracy for the early detection and classification of KOA for all four grades of KL.

Unsupervised feature learning-based encoder and adversarial networks

In this paper, we propose a novel deep learning-based feature learning architecture for object classification. Conventionally, deep learning methods are trained with supervised learning for object classification. But, this would require large amount of training data. Currently there are increasing trends to employ unsupervised learning for deep learning. By doing so, dependency on the availability of large training data could be reduced. One implementation of unsupervised deep learning is for feature learning where the network is designed to “learn” features automatically from data to obtain good representation that then could be used for classification. Autoencoder and generative adversarial networks (GAN) are examples of unsupervised deep learning methods. For GAN however, the trajectories of feature learning may go to unpredicted directions due to random initialization, making it unsuitable for feature learning. To overcome this, a hybrid of encoder and deep convolutional generative adversarial network (DCGAN) architectures, a variant of GAN, are proposed. Encoder is put on top of the Generator networks of GAN to avoid random initialisation. We called our method as EGAN. The output of EGAN is used as features for two deep convolutional neural networks (DCNNs): AlexNet and DenseNet. We evaluate the proposed methods on three types of dataset and the results indicate that better performances are achieved by our proposed method compared to using autoencoder and GAN.

반도체 불량원인 분석을 위한 딥뉴럴네트워크 기반의 패치 이미지 병합 시스템

In the integrated circuit/chip manufacturing process, failure analysis performed to find defects utilizes high-resolution chip images obtained through auto-shot scope equipment, which combines microscopy and automatic photography. However, due to the incorrect focus and the unexpected overlap size depending on the distance between the microscope and the chip, these systems are noisy. Thus, failure analysis cannot be performed effectively because the individual conduction the examination is exposed to noisy images, thereby taking a long time. We proposed a system called DeepMerge that utilizes deep learning-based learning-based features such as pint extraction and feature matching to overcome the aforementioned challenges. We will be indicating the effectiveness and efficiency of our system by obtaining practical image data from the industry.

Time series remote sensing image classification framework using combination of deep learning and multiple classifiers system

Recently, time series image (TSI) has been reported to be an effective resource to mapping fine land use/land cover (LULC), and deep learning, in particular, has been gaining growing attention in this field. However, deep learning methods using single classifier need further improvement for accurate TSI classification owing to the 1D temporal properties and insufficient dense time series of the remote sensing images. To overcome such disadvantages, we proposed an innovative approach involving construction of TSI and combination of deep learning and multiple classifiers system (MCS). Firstly, we used a normalised difference index (NDI) to establish an NDIs-based TSI and then designed a framework consisting of a deep learning-based feature extractor and multiple classifiers system (MCS) based classification model to classify the TSI. With the new approach, our experiments were conducted on Landsat images located in two counties, Sutter and Kings in California, United States. The experimental results indicate that our proposed method achieves great progress on accuracy improvement and LULC mapping, outperforming classifications using comparative deep learning and non-deep learning methods.

Deep learning-based gene selection in comprehensive gene analysis in pancreatic cancer

The selection of genes that are important for obtaining gene expression data is challenging. Here, we developed a deep learning-based feature selection method suitable for gene selection. Our novel deep learning model includes an additional feature-selection layer. After model training, the units in this layer with high weights correspond to the genes that worked effectively in the processing of the networks. Cancer tissue samples and adjacent normal pancreatic tissue samples were collected from 13 patients with pancreatic ductal adenocarcinoma during surgery and subsequently frozen. After processing, gene expression data were extracted from the specimens using RNA sequencing. Task 1 for the model training was to discriminate between cancerous and normal pancreatic tissue in six patients. Task 2 was to discriminate between patients with pancreatic cancer (n = 13) who survived for more than one year after surgery. The most frequently selected genes were ACACB, ADAMTS6, NCAM1, and CADPS in Task 1, and CD1D, PLA2G16, DACH1, and SOWAHA in Task 2. According to The Cancer Genome Atlas dataset, these genes are all prognostic factors for pancreatic cancer. Thus, the feasibility of using our deep learning-based method for the selection of genes associated with pancreatic cancer development and prognosis was confirmed.

Comparative mapping of crawling-cell morphodynamics in deep learning-based feature space

Navigation of fast migrating cells such as amoeba Dictyostelium and immune cells are tightly associated with their morphologies that range from steady polarized forms that support high directionality to those more complex and variable when making frequent turns. Model simulations are essential for quantitative understanding of these features and their origins, however systematic comparisons with real data are underdeveloped. Here, by employing deep-learning-based feature extraction combined with phase-field modeling framework, we show that a low dimensional feature space for 2D migrating cell morphologies obtained from the shape stereotype of keratocytes, Dictyostelium and neutrophils can be fully mapped by an interlinked signaling network of cell-polarization and protrusion dynamics. Our analysis links the data-driven shape analysis to the underlying causalities by identifying key parameters critical for migratory morphologies both normal and aberrant under genetic and pharmacological perturbations. The results underscore the importance of deciphering self-organizing states and their interplay when characterizing morphological phenotypes.

Towards open-set touchless palmprint recognition via weight-based meta metric learning

Touchless biometrics has become significant in the wake of novel coronavirus 2019 (COVID-19). Due to the convenience, user-friendly, and high-accuracy, touchless palmprint recognition shows great potential when the hygiene issues are considered during COVID-19. However, previous palmprint recognition methods are mainly focused on close-set scenario. In this paper, a novel Weight-based Meta Metric Learning (W2ML) method is proposed for accurate open-set touchless palmprint recognition, where only a part of categories is seen during training. Deep metric learning-based feature extractor is learned in a meta way to improve the generalization ability. Multiple sets are sampled randomly to define support and query sets, which are further combined into meta sets to constrain the set-based distances. Particularly, hard sample mining and weighting are adopted to select informative meta sets to improve the efficiency. Finally, embeddings with obvious inter-class and intra-class differences are obtained as features for palmprint identification and verification. Experiments are conducted on four palmprint benchmarks including fourteen constrained and unconstrained palmprint datasets. The results show that our W2ML method is more robust and efficient in dealing with open-set palmprint recognition issue as compared to the state-of-the-arts, where the accuracy is increased by up to 9.11% and the Equal Error Rate (EER) is decreased by up to 2.97%.

MRI‐based radiomics analysis for predicting the EGFR mutation based on thoracic spinal metastases in lung adenocarcinoma patients

This study aims to develop and evaluate multi-parametric MRI-based radiomics for preoperative identification of epidermal growth factor receptor (EGFR) mutation, which is important in treatment planning for patients with thoracic spinal metastases from primary lung adenocarcinoma. A total of 110 patients were enrolled between January 2016 and March 2019 as a primary cohort. A time-independent validation cohort was conducted containing 52 patients consecutively enrolled from July 2019 to April 2021. The patients were pathologically diagnosed with thoracic spinal metastases from primary lung adenocarcinoma; all underwent T1-weighted (T1W), T2-weighted (T2W), and T2-weighted fat-suppressed (T2FS) MRI scans of the thoracic spinal. Handcrafted and deep learning-based features were extracted and selected from each MRI modality, and used to build the radiomics signature. Various machine learning classifiers were developed and compared. A clinical-radiomics nomogram integrating the combined rad signature and the most important clinical factor was constructed with receiver operating characteristic (ROC), calibration, and decision curves analysis (DCA) to evaluate the prediction performance. The combined radiomics signature derived from the joint of three modalities can effectively classify EGFR mutation and EGFR wild-type patients, with an area under the ROC curve (AUC) of 0.886 (95% confidence interval [CI]: 0.826-0.947, SEN=0.935, SPE=0.688) in the training group and 0.803 (95% CI: 0.682-0.924, SEN=0.700, SPE=0.818) in the time-independent validation group. The nomogram incorporating the combined radiomics signature and smoking status achieved the best prediction performance in the training (AUC=0.888, 95% CI: 0.849-0.958, SEN=0.839, SPE=0.792) and time-independent validation (AUC=0.821, 95% CI: 0.692-0.929, SEN=0.667, SPE=0.909) cohorts. The DCA confirmed potential clinical usefulness of our nomogram. Our study demonstrated the potential of multi-parametric MRI-based radiomics on preoperatively predicting the EGFR mutation. The proposed nomogram model can be considered as a new biomarker to guide the selection of individual treatment strategies for patients with thoracic spinal metastases from primary lung adenocarcinoma.

AniAMPpred: artificial intelligence guided discovery of novel antimicrobial peptides in animal kingdom.

With advancements in genomics, there has been substantial reduction in the cost and time of genome sequencing and has resulted in lot of data in genome databases. Antimicrobial host defense proteins provide protection against invading microbes. But confirming the antimicrobial function of host proteins by wet-lab experiments is expensive and time consuming. Therefore, there is a need to develop an in silico tool to identify the antimicrobial function of proteins. In the current study, we developed a model AniAMPpred by considering all the available antimicrobial peptides (AMPs) of length $\in $[10 200] from the animal kingdom. The model utilizes a support vector machine algorithm with deep learning-based features and identifies probable antimicrobial proteins (PAPs) in the genome of animals. The results show that our proposed model outperforms other state-of-the-art classifiers, has very high confidence in its predictions, is not biased and can classify both AMPs and non-AMPs for a diverse peptide length with high accuracy. By utilizing AniAMPpred, we identified 436 PAPs in the genome of Helobdella robusta. To further confirm the functional activity of PAPs, we performed BLAST analysis against known AMPs. On detailed analysis of five selected PAPs, we could observe their similarity with antimicrobial proteins of several animal species. Thus, our proposed model can help the researchers identify PAPs in the genome of animals and provide insight into the functional identity of different proteins. An online prediction server is also developed based on the proposed approach, which is freely accessible at https://aniamppred.anvil.app/.

Deep learning-based meta-classifier approach for COVID-19 classification using CT scan and chest X-ray images.

Literature survey shows that convolutional neural network (CNN)-based pretrained models have been largely used for CoronaVirus Disease 2019 (COVID-19) classification using chest X-ray (CXR) and computed tomography (CT) datasets. However, most of the methods have used a smaller number of data samples for both CT and CXR datasets for training, validation, and testing. As a result, the model might have shown good performance during testing, but this type of model will not be more effective on unseen COVID-19 data samples. Generalization is an important term to be considered while designing a classifier that can perform well on completely unseen datasets. Here, this work proposes a large-scale learning with stacked ensemble meta-classifier and deep learning-based feature fusion approach for COVID-19 classification. The features from the penultimate layer (global average pooling) of EfficientNet-based pretrained models were extracted and the dimensionality of the extracted features reduced using kernel principal component analysis (PCA). Next, a feature fusion approach was employed to merge the features of various extracted features. Finally, a stacked ensemble meta-classifier-based approach was used for classification. It is a two-stage approach. In the first stage, random forest and support vector machine (SVM) were applied for prediction, then aggregated and fed into the second stage. The second stage includes logistic regression classifier that classifies the data sample of CT and CXR into either COVID-19 or Non-COVID-19. The proposed model was tested using large CT and CXR datasets, which are publicly available. The performance of the proposed model was compared with various existing CNN-based pretrained models. The proposed model outperformed the existing methods and can be used as a tool for point-of-care diagnosis by healthcare professionals.

Automatic Detection of Coronavirus Disease (COVID-19) in X-ray and CT Images: A Machine Learning Based Approach

The newly identified Coronavirus pneumonia, subsequently termed COVID-19, is highly transmittable and pathogenic with no clinically approved antiviral drug or vaccine available for treatment. The most common symptoms of COVID-19 are dry cough, sore throat, and fever. Symptoms can progress to a severe form of pneumonia with critical complications, including septic shock, pulmonary edema, acute respiratory distress syndrome and multi-organ failure. While medical imaging is not currently recommended in Canada for primary diagnosis of COVID-19, computer-aided diagnosis systems could assist in the early detection of COVID-19 abnormalities and help to monitor the progression of the disease, potentially reduce mortality rates. In this study, we compare popular deep learning-based feature extraction frameworks for automatic COVID-19 classification. To obtain the most accurate feature, which is an essential component of learning, MobileNet, DenseNet, Xception, ResNet, InceptionV3, InceptionResNetV2, VGGNet, NASNet were chosen amongst a pool of deep convolutional neural networks. The extracted features were then fed into several machine learning classifiers to classify subjects as either a case of COVID-19 or a control. This approach avoided task-specific data pre-processing methods to support a better generalization ability for unseen data. The performance of the proposed method was validated on a publicly available COVID-19 dataset of chest X-ray and CT images. The DenseNet121 feature extractor with Bagging tree classifier achieved the best performance with 99% classification accuracy. The second-best learner was a hybrid of the a ResNet50 feature extractor trained by LightGBM with an accuracy of 98