Convolutional Neural Network Transfer Learning Research Articles

Deep COVID 19: Deep Learning for COVID 19 Detection from X ray Images

The COVID-19 will take place for the first time in December 2019 in Wuhan, China. After that, the virus spread all over the world, with over 4.7 million confirmed cases and over 315000 deaths as of the time of writing this report. Radiologists can employ machine learning algorithms developed on radiography pictures as a decision support mechanism to help them speed up the diagnostic process. The goal of this study is to conduct a quantitative evaluation of six off-the-shelf convolutional neural networks (CNNs) for COVID-19 X-ray image analysis. Due to the limited amount of images available for analysis, the CNN transfer learning approach was used. We also developed a simple CNN architecture with a modest number of parameters that does a good job of differentiating COVID-19 from regular X-rays. in this paper, we are used large dataset which contained CXR images of normal patients and patients with COVID-19. the number of CXR images for normal patients are 10,192 image and the number of CXR images for COVID-19 patients are 3,616 images. The results of experiments show the effectiveness and robustness of Deep-COVID-19 and pretrained models like VGG16, VGG19, and MobileNets. Our proposed Model Deep-COVID-19 achieved over 94.5% accuracy.

Domain- and task-specific transfer learning for medical segmentation tasks

Background and objectives: Transfer learning is a valuable approach to perform medical image segmentation in settings with limited cases available for training convolutional neural networks (CNN). Both the source task and the source domain influence transfer learning performance on a given target medical image segmentation task. This study aims to assess transfer learning-based medical segmentation task performance for various source task and domain combinations. Methods: CNNs were pre-trained on classification, segmentation, and self-supervised tasks on two domains: natural images and T1 brain MRI. Next, these CNNs were fine-tuned on three target T1 brain MRI segmentation tasks: stroke lesion, MS lesions, and brain anatomy segmentation. In all experiments, the CNN architecture and transfer learning strategy were the same. The segmentation accuracy on all target tasks was evaluated using the mIOU or Dice coefficients. The detection accuracy was evaluated for the stroke and MS lesion target tasks only. Results: CNNs pre-trained on a segmentation task on the same domain as the target tasks resulted in higher or similar segmentation accuracy compared to other source task and domain combinations. Pre-training a CNN on ImageNet resulted in a comparable, but not consistently higher lesion detection rate, despite the amount of training data used being 10 times larger. Conclusions: This study suggests that optimal transfer learning for medical segmentation is achieved with a similar task and domain for pre-training. As a result, CNNs can be effectively pre-trained on smaller datasets by selecting a source domain and task similar to the target domain and task.

Crackle Detection In Lung Sounds Using Transfer Learning And Multi-Input Convolutional Neural Networks.

Large annotated lung sound databases are publicly available and might be used to train algorithms for diagnosis systems. However, it might be a challenge to develop a well-performing algorithm for small non-public data, which have only a few subjects and show differences in recording devices and setup. In this paper, we use transfer learning to tackle the mismatch of the recording setup. This allows us to transfer knowledge from one dataset to another dataset for crackle detection in lung sounds. In particular, a single input convolutional neural network (CNN) model is pre-trained on a source domain using ICBHI 2017, the largest publicly available database of lung sounds. We use log-mel spectrogram features of respiratory cycles of lung sounds. The pre-trained network is used to build a multi-input CNN model, which shares the same network architecture for respiratory cycles and their corresponding respiratory phases. The multi-input model is then fine-tuned on the target domain of our self-collected lung sound database for classifying crackles and normal lung sounds. Our experimental results show significant performance improvements of 9.84% (absolute) in F-score on the target domain using the multi-input CNN model and transfer learning for crackle detection.Clinical relevance- Crackle detection in lung sounds, multi-input convolutional neural networks, transfer learning.

White matter structural connectivity as a biomarker for detecting juvenile myoclonic epilepsy by transferred deep convolutional neural networks with varying transfer rates

Objective. By detecting abnormal white matter changes, diffusion magnetic resonance imaging (MRI) contributes to the detection of juvenile myoclonic epilepsy (JME). In addition, deep learning has greatly improved the detection performance of various brain disorders. However, there is almost no previous study effectively detecting JME by a deep learning approach with diffusion MRI. Approach. In this study, the white matter structural connectivity was generated by tracking the white matter fibers in detail based on Q-ball imaging and neurite orientation dispersion and density imaging. Four advanced deep convolutional neural networks (CNNs) were deployed by using the transfer learning approach, in which the transfer rate searching strategy was proposed to achieve the best detection performance. Main results. Our results showed: (a) Compared to normal control, the white matter’ neurite density of JME was significantly decreased. The most significantly abnormal fiber tracts between the two groups were found to be cortico-cortical connection tracts. (b) The proposed transfer rate searching approach contributed to find each CNN’s best performance, in which the best JME detection accuracy of 92.2% was achieved by using the Inception_resnet_v2 network with a 16% transfer rate. Significance. The results revealed: (a) Through detection of the abnormal white matter changes, the white matter structural connectivity can be used as a useful biomarker for detecting JME, which helps to characterize the pathophysiology of epilepsy. (b) The proposed transfer rate, as a new hyperparameter, promotes the CNNs transfer learning performance in detecting JME.

A novel 1DCNN and domain adversarial transfer strategy for small sample GIS partial discharge pattern recognition

Recently, convolutional neural networks (CNNs) have made certain achievements in gas-insulated switchgear (GIS) partial discharge (PD) pattern recognition. However, these methods rely on the availability of massive PD samples and how to apply the CNN constructed in the laboratory to the field GIS PD pattern recognition has become an urgent problem. To solve these problems, we propose a small sample GIS PD pattern recognition using one-dimensional CNN (1DCNN) and domain adversarial transfer learning (DATL). First, a novel 1DCNN is constructed to achieve high-accuracy classification using unbalanced samples, where the problem of traditional two-dimensional CNN’s dependence on sample size is solved. Second, DATL is used to realize on-site GIS PD pattern recognition using small samples containing some unlabeled samples. In the domain adversarial training, two domain classifiers are introduced to align the domain of the decision boundary, which achieves a suitable features migration and accurate classification of target domains. Through the construction of multiple experiments, we verified that the proposed method achieves 98.67% and >92% recognition accuracy in the source domain and target domain, respectively. Compared with the existing methods, the proposed method can realize satisfactory pattern recognition, which can provide strong support for the subsequent pattern recognition of GIS PD.

Cotton-wool spots, red-lesions and hard-exudates distinction using CNN enhancement and transfer learning

<div>The automatic retinal disease diagnosis by artificial intelligent is an interesting and challenging topic in the medical field. It requires an appropriate image enhancement technique and a sufficient training dataset for the specific retina conditions. The aim of this study was to design an automatic diagnosis convolutional neural network (CNN) model which does not require a large training dataset to specifically identify diabetic retinopathy symptoms, which are cotton wool, exudates spots and red lesionin colour fundus pictures. A novel framework comprised image enhancement method by using upgraded contrast limited adaptive histogram equalization (UCLAHE) filter and transferred pre-trained networks was developed to classify the retinal diseases regarding to the symptoms. The performance of the proposed framework was evaluated based on accuracy, sensitivity and specificity metrics. The collected results have proven the robustness of the proposed framework in offering good accuracy in retina diseases diagnosis. </div>

An Analysis on Object Recognition Using Convolutional Neural Networks

The global development and progress in scientific paraphernalia and technology is the fundamental reason for the rapid increasein the data volume. Several significant techniques have been introducedfor image processing and object detection owing to this advancement. The promising features and transfer learning of ConvolutionalNeural Network (CNN) havegained much attention around the globe by researchers as well as computer vision society, as a result of which, several remarkable breakthroughs were achieved. This paper comprehensively reviews the data classification, history as well as architecture of CNN and well-known techniques bytheir boons and absurdities. Finally, a discussion for implementation of CNN over object detection for effectual results based on their critical analysis and performances is presented

Classification of Healthy Control and Abnormal Lung Chest Radiography images using CBIR and Atlas-Based Graph cut Segmentation by Transfer Learning CNNs

The prominent aim of systematic lung pathologies detection is to improve health outcomes among people with pulmonary diseases and decrease thorax diseases disseminating among the peoples through improved lung abnormality detection, reduction in diagnostic delays and early treatment. Today technological breakthroughs have led to the discovery of various invasive clinical imaging methods for the analysis of lung diseases. In this work, an integrated Computer-Aided Diagnoses (CAD) based on Deep Learning (DL) to diagnosis of lung diseases on chest radiographs was proposed. In this paper we have used JSRT datasets it contains 247 images are employed atlas-based graph-cut segmentation algorithm for extract the lung regions and segmented regions are input to the Convolutional Neural Network (CNN) with augmentation techniques that are provides the training accuracy is 97% and 89% for testing accuracy.

Partial Discharge Online Detection for Long-Term Operational Sustainability of On-Site Low Voltage Distribution Network Using CNN Transfer Learning

Partial discharge (PD) detection studies aiming at the fault diagnosis for facilities and power cables in transmission networks have been conducted over the years. Recently, the deep learning models for PD detection have been used to diagnose the PD fault of facilities and cables. Most PD studies have been conducted in the field, such as gas-insulated switchgear (GIS) and power cables for high voltage transmission networks. There are few studies of PD fault detection for on-site low-voltage distribution networks. Additionally, there are few studies of PD detection algorithms for improving the accuracy of the deep learning models using small real PD data only. In this study, a PD online detection system and a model for long-term operational sustainability of on-site low voltage distribution networks are proposed using convolutional neural network (CNN) transfer-learning. The proposed PD online system makes it possible to acquire as many real PD data as possible through continuous monitoring of PD occurrence. The PD detection accuracy results showed that the proposed CNN transfer-learning models are more effective models for obtaining improved accuracy (97.4%) than benchmark models, such as CNN and support vector machine (SVM) using only small real PD data acquired from PD online detection system.

Deep Transfer Learning for Automated Intestinal Bleeding Detection in Capsule Endoscopy Imaging.

The objective of this paper was to develop a computer-aided diagnostic (CAD) tools for automated analysis of capsule endoscopic (CE) images, more precisely, detect small intestinal abnormalities like bleeding. In particular, we explore a convolutional neural network (CNN)-based deep learning framework to identify bleeding and non-bleeding CE images, where a pre-trained AlexNet neural network is used to train a transfer learning CNN that carries out the identification. Moreover, bleeding zones in a bleeding-identified image are also delineated using deep learning-based semantic segmentation that leverages a SegNet deep neural network. To evaluate the performance of the proposed framework, we carry out experiments on two publicly available clinical datasets and achieve a 98.49% and 88.39% F1 score, respectively, on the capsule endoscopy.org and KID datasets. For bleeding zone identification, 94.42% global accuracy and 90.69% weighted intersection over union (IoU) are achieved. Finally, our performance results are compared to other recently developed state-of-the-art methods, and consistent performance advances are demonstrated in terms of performance measures for bleeding image and bleeding zone detection. Relative to the present and established practice of manual inspection and annotation of CE images by a physician, our framework enables considerable annotation time and human labor savings in bleeding detection in CE images, while providing the additional benefits of bleeding zone delineation and increased detection accuracy. Moreover, the overall cost of CE enabled by our framework will also be much lower due to the reduction of manual labor, which can make CE affordable for a larger population.

Fault Diagnosis of Reciprocating Compressor Valve Based on Transfer Learning Convolutional Neural Network

Reciprocating compressors play a vital role in oil, natural gas, and general industrial processes. Their safe and stable operation directly affects the healthy development of the enterprise economy. Since the valve failure accounts for 60% of the total failures when the reciprocating compressor fails, it is of great significance to quickly find and diagnose the failure type of the valve for the fault diagnosis of the reciprocating compressor. At present, reciprocating compressor valve fault diagnosis based on deep neural networks requires sufficient labeled data for training, but valve in real-case reciprocating compressor (VRRC) does not have enough labeled data to train a reliable model. Fortunately, the data of valve in laboratory reciprocating compressor (VLRC) contains relevant fault diagnosis knowledge. Therefore, inspired by the idea of transfer learning, a fault diagnosis method for reciprocating compressor valves based on transfer learning convolutional neural network (TCNN) is proposed. This method uses convolutional neural network (CNN) to extract the transferable features of gas temperature and pressure data from VLRC and VRRC and establish pseudolabels for VRRC unlabeled data. Three regularization terms, the maximum mean discrepancy (MMD) of the transferable features of VLRC and VRRC data, the error between the VLRC sample label prediction and the actual label, and the error between the VRRC sample label prediction and the pseudolabel, are proposed. Their weighted sum is used as an objective function to train the model, thereby reducing the distribution difference of domain feature transfer and increasing the distance between learning feature classes. Experimental results show that this method uses VLRC data to identify the health status of VRRC, and the fault recognition rate can reach 98.32%. Compared with existing methods, this method has higher diagnostic accuracy, which proves the effectiveness of this method.

DLAM: Deep Learning Based Real-Time Porosity Prediction for Additive Manufacturing Using Thermal Images of the Melt Pool

This paper presents an investigation of the rapid variations in the temperature of metal melt pool for Additive Manufacturing (AM) processes. The melt pool is created by scanning a high-power laser beam across a metal powder bed. Rapid heating and cooling processes are involved in the layer-by-layer fabrication of the metal part. Recent advances in Machine Learning and Deep Learning algorithms provide efficient ways to analyze large sets of data in search of correlations that would otherwise be extremely time-consuming. The use of Machine Learning and Deep Learning algorithms to understand temperature variations in AM fabrication process will allow to predict the formation of porosity before it occurs. The objective of this research is to advance the AM technology using enhanced Deep Learning techniques to provide in-situ analysis of the melt pool temperature that can lead to a reliable manufacturing of Three-Dimensional (3D) metal parts/components. In specific, Deep Learning based porosity prediction for Additive Manufacturing (DLAM) methods have been proposed. In DLAMs, several state-of-the-art Deep Learning algorithms such as Convolutional Neural Networks (CNN) using transfer learning, and Residual-Recurrent Convolutional Neural Networks (Res-RCNN) are proposed for effectively performing the end-to-end porosity prediction in real-time using thermal images of melt pool. Experimental results, in this research, show that the Res-RCNN has an overall accuracy of 99.49% and inference time of $8.67ms$ , and the Res-RCNN outperforms other baseline models. The Res-RCNN’s recursive architecture allows the network to view each input image multiples times and at varying feature levels, which enables a slight boost in porosity prediction accuracy over the commonly used transfer learning CNN models.

Automated detection of multiple structural changes of diabetic macular oedema in SDOCT retinal images through transfer learning in CNNs

Diabetic Maculopathy, the major cause of vision loss, occurs due to uncontrolled diabetes. This affects the retinal layers of the eye causing bleeding of vessels, which results in Diabetic Macular Edema (DME), macular detachment etc., Basically three structural changes are involved in DME viz., Cystoid Macular Edema (CME), Serous Macular Detachment (SMD) and Intra Retinal Fluid (IRF). These changes may also coexist with each other such as CME with SMD or CME with IRF etc. In this work, the retinal images acquired through Spectral Domain Optical Coherence Tomography (SDOCT) imaging modality, which would provide high level of precision and resolution of the retinal layers are utilized. An automated algorithm to detect and differentiate seven types of DME based on deep learning is implemented using transfer learning on three Convolutional Neural Networks (CNNs), ResNet 50, VGGNet and AlexNet. In detection of DME, ResNet 50 performs excellently well, when compared with Alexnet and VGGNet, because of its depth and skip connection features. The average values of the statistical parameters such as accuracy (0.993), F1 score (0.975), Mathews Correlation Coefficient(MCC)(0.972) of ResNet are high when compared to that of AlexNet and VGGNet

Multimodal biometric recognition systems using deep learning based on the finger vein and finger knuckle print fusion

Recognition systems using multimodal biometrics attracts attention because they improve recognition efficiency and high-security level compared to the unimodal biometrics system. In this study, the authors present a secure multimodal biometrics recognition system based on the deep learning method that uses convolutional neural networks (CNNs). The authors propose two multimodal architectures using the finger knuckle print (FKP) and the finger vein (FV) biometrics with different levels of fusion: the features level fusion and scores level fusion. The features extraction for FKP and FV are performed using transfer learning CNN architectures: AlexNet, VGG16, and ResNet50. The key step aims to select separate features descriptors from each unimodal biometrics modality. After that, the authors combine them using the proposed fusion approaches were support vector machine or Softmax applies as classifiers to increase the proposed system security. The efficiency of the proposed algorithms is tested using publicly available biometrics databases. The experimental results show that the proposed fusion architectures achieve an accuracy of 99.89% and an equal error rate of 0.05%. The obtained results indicate that the biometrics recognition system using deep learning is secure, robust, and reliable.

Deep Transfer Learning for Vulnerable Road Users Detection using Smartphone Sensors Data

As the Autonomous Vehicle (AV) industry is rapidly advancing, the classification of non-motorized (vulnerable) road users (VRUs) becomes essential to ensure their safety and to smooth operation of road applications. The typical practice of non-motorized road users’ classification usually takes significant training time and ignores the temporal evolution and behavior of the signal. In this research effort, we attempt to detect VRUs with high accuracy be proposing a novel framework that includes using Deep Transfer Learning, which saves training time and cost, to classify images constructed from Recurrence Quantification Analysis (RQA) that reflect the temporal dynamics and behavior of the signal. Recurrence Plots (RPs) were constructed from low-power smartphone sensors without using GPS data. The resulted RPs were used as inputs for different pre-trained Convolutional Neural Network (CNN) classifiers including constructing 227 × 227 images to be used for AlexNet and SqueezeNet; and constructing 224 × 224 images to be used for VGG16 and VGG19. Results show that the classification accuracy of Convolutional Neural Network Transfer Learning (CNN-TL) reaches 98.70%, 98.62%, 98.71%, and 98.71% for AlexNet, SqueezeNet, VGG16, and VGG19, respectively. Moreover, we trained resnet101 and shufflenet for a very short time using one epoch of data and then used them as weak learners, which yielded 98.49% classification accuracy. The results of the proposed framework outperform other results in the literature (to the best of our knowledge) and show that using CNN-TL is promising for VRUs classification. Because of its relative straightforwardness, ability to be generalized and transferred, and potential high accuracy, we anticipate that this framework might be able to solve various problems related to signal classification.

Issues associated with deploying CNN transfer learning to detect COVID-19 from chest X-rays.

Covid-19 first occurred in Wuhan, China in December 2019. Subsequently, the virus spread throughout the world and as of June 2020 the total number of confirmed cases are above 4.7 million with over 315,000 deaths. Machine learning algorithms built on radiography images can be used as a decision support mechanism to aid radiologists to speed up the diagnostic process. The aim of this work is to conduct a critical analysis to investigate the applicability of convolutional neural networks (CNNs) for the purpose of COVID-19 detection in chest X-ray images and highlight the issues of using CNN directly on the whole image. To accomplish this task, we use 12-off-the-shelf CNN architectures in transfer learning mode on 3 publicly available chest X-ray databases together with proposing a shallow CNN architecture in which we train it from scratch. Chest X-ray images are fed into CNN models without any preprocessing to replicate researches used chest X-rays in this manner. Then a qualitative investigation performed to inspect the decisions made by CNNs using a technique known as class activation maps (CAM). Using CAMs, one can map the activations contributed to the decision of CNNs back to the original image to visualize the most discriminating region(s) on the input image. We conclude that CNN decisions should not be taken into consideration, despite their high classification accuracy, until clinicians can visually inspect and approve the region(s) of the input image used by CNNs that lead to its prediction.

Convolutional Neural Network Transfer Learning Applied to the Affective Auditory P300-Based BCI

Brain-computer interface (BCI) enables us to interact with the external world via electroencephalography (EEG) signals. Recently, deep learning methods have been applied to the BCI to reduce the time required for recording training data. However, more evidence is required due to lack of comparison. To reveal more evidence, this study proposed a deep learning method named time-wise convolutional neural network (TWCNN), which was applied to a BCI dataset. In the evaluation, EEG data from a subject was classified utilizing previously recorded EEG data from other subjects. As a result, TWCNN showed the highest accuracy, which was significantly higher than the typically used classifier. The results suggest that the deep learning method may be useful to reduce the recording time of training data.

The Optimization of Transfer Learning Convolutional Neural Network Model with PCA and t-SNE Algorithms for Classification and Recognition of CIFAR-10 Image

The low accuracy when performing the image classification process is a problem that often occurs. The image classification process requires the completeness of the features of the image which form an informative image pattern so that information from the image can be displayed. The purpose of this study is to classify images in the CIFAR-10 image dataset using the CNN method. Initially the CNN method gave an accuracy of 79.4% but had a long computation time of 12 hours with 10,000 iterations. The optimization process for the CNN method is carried out by combining the CNN method, the PCA algorithm and the t-SNE algorithm. The algorithm is used to reduce the length of the image matrix in the initial transfer of learning without reducing the information in the image so that the classification process can be done correctly. The final result obtained from the optimization has an accuracy of 90.5%. With an optimization rate of 11%. The resulting time is more efficient, namely 3 hours for the feature transfer-value process and 6 minutes for the testing process with 10,000 iterations.

Detection of Mesangial hypercellularity of MEST-C score in immunoglobulin A-nephropathy using deep convolutional neural network

Immunoglobulin A (IgA)-nephropathy (IgAN) is one of the major reasons for renal failure. It provides vital clues to estimate the stage and the proliferation rate of end-stage kidney disease. IgA stage can be estimated with the help of MEST-C score. The manual estimation of MEST-C score from whole slide kidney images is a very tedious and difficult task. This study uses some Convolutional neural networks (CNNs) related models to detect mesangial hypercellularity (M score) in MEST-C. CNN learns the features directly from image data without the requirement of analytical data. CNN is trained efficiently when image data size is large enough for a particular class. In the case of smaller data size, transfer learning can be used efficiently in which CNN is pre-trained on some general images and then on subject images. Since the data set size is small, time spent in collecting large data set is saved. The training time of transfer learning is also reduced because the model is already pre-trained. This research work aims at the detection of mesangial hypercellularity from biopsy images with small data size by utilizing the transfer learning. The dataset used in this research work consists of 138 individual glomerulus (× 20 magnification digital biopsy) images of IgA patients received from All India Institute of Medical Science, Delhi. Here, machine learning (k-nearest neighbour (KNN) and support vector machine (SVM)) classifiers are compared to transfer learning CNN methods. The deep extracted image features are used by machine learning classifiers. The different evaluation parameters have been used for comparing the predictions of basic classifiers to the deep learning model. The research work concludes that the transfer learning deep CNN method can improve the detection of mesangial hypercellularity as compare to KNN, SVM methods when using the small data set. This model could help the pathologists to understand the stages of kidney failure.

Transfer learning using a multi-scale and multi-network ensemble for skin lesion classification

Background and objective: Skin cancer is among the most common cancer types in the white population and consequently computer aided methods for skin lesion classification based on dermoscopic images are of great interest. A promising approach for this uses transfer learning to adapt pre-trained convolutional neural networks (CNNs) for skin lesion diagnosis. Since pre-training commonly occurs with natural images of a fixed image resolution and these training images are usually significantly smaller than dermoscopic images, downsampling or cropping of skin lesion images is required. This however may result in a loss of useful medical information, while the ideal resizing or cropping factor of dermoscopic images for the fine-tuning process remains unknown.Methods: We investigate the effect of image size for skin lesion classification based on pre-trained CNNs and transfer learning. Dermoscopic images from the International Skin Imaging Collaboration (ISIC) skin lesion classification challenge datasets are either resized to or cropped at six different sizes ranging from 224 × 224 to 450 × 450. The resulting classification performance of three well established CNNs, namely EfficientNetB0, EfficientNetB1 and SeReNeXt-50 is explored. We also propose and evaluate a multi-scale multi-CNN (MSM-CNN) fusion approach based on a three-level ensemble strategy that utilises the three network architectures trained on cropped dermoscopic images of various scales.Results: Our results show that image cropping is a better strategy compared to image resizing delivering superior classification performance at all explored image scales. Moreover, fusing the results of all three fine-tuned networks using cropped images at all six scales in the proposed MSM-CNN approach boosts the classification performance compared to a single network or a single image scale. On the ISIC 2018 skin lesion classification challenge test set, our MSM-CNN algorithm yields a balanced multi-class accuracy of 86.2% making it the currently second ranked algorithm on the live leaderboard.Conclusions: We confirm that the image size has an effect on skin lesion classification performance when employing transfer learning of CNNs. We also show that image cropping results in better performance compared to image resizing. Finally, a straightforward ensembling approach that fuses the results from images cropped at six scales and three fine-tuned CNNs is shown to lead to the best classification performance.