Proposed CNN Model Research Articles

Detecting Internet of Things attacks using distributed deep learning

The reliability of Internet of Things (IoT) connected devices is heavily dependent on the security model employed to protect user data and prevent devices from engaging in malicious activity. Existing approaches for detecting phishing, distributed denial of service (DDoS), and Botnet attacks often focus on either the device or the back-end. In this paper, we propose a cloud-based distributed deep learning framework for phishing and Botnet attack detection and mitigation. The model comprises two key security mechanisms working cooperatively, namely: (1) a Distributed Convolutional Neural Network (DCNN) model embedded as an IoT device micro-security add-on for detecting phishing and application layer DDoS attacks; and (2) a cloud-based temporal Long-Short Term Memory (LSTM) network model hosted on the back-end for detecting Botnet attacks, and ingest CNN embeddings to detect distributed phishing attacks across multiple IoT devices. The distributed CNN model, embedded into a ML engine in the client's IoT device, allows us to detect and defend the IoT device from phishing attacks at the point of origin. We create a dataset consisting of both phishing and non-phishing URLs to train the proposed CNN add-on security model, and select the N_BaIoT dataset for training the back-end LSTM model. The joint training method minimizes communication and resource requirements for attack detection, and maximizes the usefulness of extracted features. In addition, an aggregation of schemes allows the automatic fusion of multiple requests to improve the overall performance of the system. Our experiments show that the IoT micro-security add-on running the proposed CNN model is capable of detecting phishing attacks with an accuracy of 94.3% and a F-1 score of 93.58%. Using the back-end LSTM model, the model detects Botnet attacks with an accuracy of 94.80% using all malicious data points in the used dataset. Thus, the findings demonstrate that the proposed approach is capable of detecting attacks, both at device and at the back-end level, in a distributed fashion.

Asian Female Facial Beauty Prediction Using Deep Neural Networks via Transfer Learning and Multi-Channel Feature Fusion

Facial beauty plays an important role in many fields today, such as digital entertainment, facial beautification surgery and etc. However, the facial beauty prediction task has the challenges of insufficient training datasets, low performance of traditional methods, and rarely takes advantage of the feature learning of Convolutional Neural Networks. In this paper, a transfer learning based CNN method that integrates multiple channel features is utilized for Asian female facial beauty prediction tasks. Firstly, a Large-Scale Asian Female Beauty Dataset (LSAFBD) with a more reasonable distribution has been established. Secondly, in order to improve CNN’s self-learning ability of facial beauty prediction task, an effective CNN using a novel Softmax-MSE loss function and a double activation layer has been proposed. Then, a data augmentation method and transfer learning strategy were also utilized to mitigate the impact of insufficient data on proposed CNN performance. Finally, a multi-channel feature fusion method was explored to further optimize the proposed CNN model. Experimental results show that the proposed method is superior to traditional learning method combating the Asian female FBP task. Compared with other state-of-the-art CNN models, the proposed CNN model can improve the rank-1 recognition rate from 60.40% to 64.85%, and the pearson correlation coefficient from 0.8594 to 0.8829 on the LSAFBD and obtained 0.9200 regression prediction results on the SCUT dataset.

On-Road Deer Detection for Advanced Driver Assistance using Convolutional Neural Network

Animal-vehicle collision (AVC) is a major concern in road safety that affects human life, properties, and wildlife. Most of the collisions happen with large animals especially deer that enters the road suddenly. Furthermore, the threat is even more alarming in poor visibility conditions such as night-time, fog, rain, etc. Therefore, it is vital to detect the presence of deer on roadways to mitigate the severity of deer-vehicle collision (DVC). This paper presents an efficient methodology to detect deer on roadways both during the day and night-time conditions using deep learning framework. A two-class CNN model differentiating a deer from its background is developed. The background will have a few classes of objects such as motorcycles, cars, and trees which are frequently encountered on roadways. A self-constructed dataset with both RGB and thermal images is used to train the CNN model. Sliding window technique is used to localize the spatial region of deer in an image. The performance of the proposed CNN model is compared with state-of-the art classifiers and pre-trained CNN models and the results validate its effectiveness.

COVID-19 diagnosis from chest X-ray images using transfer learning: Enhanced performance by debiasing dataloader.

BACKGROUND:Chest X-ray imaging has been proved as a powerful diagnostic method to detect and diagnose COVID-19 cases due to its easy accessibility, lower cost and rapid imaging time.OBJECTIVE: This study aims to improve efficacy of screening COVID-19 infected patients using chest X-ray images with the help of a developed deep convolutional neural network model (CNN) entitled nCoV-NET. METHODS: To train and to evaluate the performance of the developed model, three datasets were collected from resources of “ChestX-ray14”, “COVID-19 image data collection”, and “Chest X-ray collection from Indiana University,” respectively. Overall, 299 COVID-19 pneumonia cases and 1,522 non-COVID 19 cases are involved in this study. To overcome the probable bias due to the unbalanced cases in two classes of the datasets, ResNet, DenseNet, and VGG architectures were re-trained in the fine-tuning stage of the process to distinguish COVID-19 classes using a transfer learning method. Lastly, the optimized final nCoV-NET model was applied to the testing dataset to verify the performance of the proposed model. RESULTS:Although the performance parameters of all re-trained architectures were determined close to each other, the final nCOV-NET model optimized by using DenseNet-161 architecture in the transfer learning stage exhibits the highest performance for classification of COVID-19 cases with the accuracy of 97.1 %. The Activation Mapping method was used to create activation maps that highlights the crucial areas of the radiograph to improve causality and intelligibility.CONCLUSION: This study demonstrated that the proposed CNN model called nCoV-NET can be utilized for reliably detecting COVID-19 cases using chest X-ray images to accelerate the triaging and save critical time for disease control as well as assisting the radiologist to validate their initial diagnosis.

Classification of Myocardial Infarction using Convolution Neural Network

Myocardial infarction is one of the most dangerous cardiovascular diseases for most of the people in the world. It is generally confessed as a heart attack. The damage of the heart muscle causes the Myocardial Infraction (MI). When there is a block in heart veins, then the flow of oxygen to the heart muscle also gets blocked, which leads to damage of the heart muscle. The damage is irreversible, so it may lead to death. Quick and exact recognition of MI is required to reduce the death rate. There are several diagnostic tools such as blood tests, ECG is available for the analysis of cardiovascular disease. Among all tools, ECG provides effective results in determining MI, but the manual interpretation of the ECG signal may take time for the doctor to identify the symptoms of MI. The manual interpretation may vary from person to person. Hence a computer-aided diagnostic tool is required to analyze ECG signals effectively for identifying MI. This paper aims to provide an algorithm for the detection of myocardial infarction that operates directly on ECG data. Nowadays Convolutional neural network is cable of analyzing an image effectively so, a deep learning model with the CNN algorithm is used in this paper to classify the images and to identify whether the image has MI or not. The proposed CNN model yields 87% accuracy for the Physikalisch-Technische Bundesanstalt database.

Temporally downsampled cerebral CT perfusion image restoration using deep residual learning.

Acute ischemic stroke is one of the most causes of death all over the world. Onset to treatment time is critical in stroke diagnosis and treatment. Considering the time consumption and high price of MR imaging, CT perfusion (CTP) imaging is strongly recommended for acute stroke. However, too much CT radiation during CTP imaging may increase the risk of health problems. How to reduce CT radiation dose in CT perfusion imaging has drawn our great attention. In this study, the original 30-pass CTP images are downsampled to 15 passes in time sequence, which equals to 50% radiation dose reduction. Then, a residual deep convolutional neural network (DCNN) model is proposed to restore the downsampled 15-pass CTP images to 30 passes to calculate the parameters such as cerebral blood flow, cerebral blood volume, mean transit time, time to peak for stroke diagnosis and treatment. The deep restoration CNN is implemented simply and effectively with 16 successive convolutional layers which form a wide enough receptive field for input image data. 18 patients' CTP images are employed as training set and the other six patients' CTP images are treated as test dataset in this study. Experiments demonstrate that our CNN can restore high-quality CTP images in terms of structural similarity index (SSIM) and peak signal-to-noise ratio (PSNR). The average SSIM and PSNR for test images are 0.981 and 56.25, and the SSIM and PSNR of regions of interest are 0.915 and 42.44, respectively, showing promising quantitative level. In addition, we compare the perfusion maps calculated from the restored images and from the original images, and the average perfusion results of them are extremely close. Areas of hypoperfusion of six test cases could be detected with comparable accuracy by radiologists. The trained model can restore the temporally downsampled 15-pass CTP to 30 passes very well. According to the contrast test, sufficient information cannot be restored with, e.g., simple interpolation method and deep convolutional generative adversarial network, but can be restored with the proposed CNN model. This method can be an optional way to reduce radiation dose during CTP imaging.

Deep convolutional neural network for fast determination of the rock strength parameters using drilling data

Accurate, fast, and reliable estimation of field rock strength is significant for the assessment of rock mass strength. This paper presents a novel method to continuously estimate the field strength parameters of rock by incorporating a deep convolutional neural network (DCNN) technique into the drilling process. A DCNN model is established based on the stochastic pooling method and softmax loss. The proposed CNN model framework includes: (1) collecting drilling performance parameters of intact rock, (2) establishing a strength parameter database via standard laboratory tests, (3) training the proposed CNN model using images created from the dataset of drilling performance parameters and rock strength parameters (unconfined compressive strength (UCS), cohesion, and internal friction angle), and (4) performing a new estimation to evaluate the reliability of the trained CNN model. The experimental results show that the predicted rock strength parameters are within the accepted error range of 10% compared to the results from the conventional standard test. The proposed CNN shows superior performance for UCS estimation of various rock types and achieves higher accuracy than the Mohr–Coulomb criterion. In particular, the proposed method can overcome major limitations of analytical models based on the force limit and energy equilibrium method. This method enables continuous and reliable field measurements of rock strength parameters at a speed several orders of magnitude faster than the standard test. This continuous and practical technique has been applied in an engineering problem in the Hanjiang-to-Weihe River Project of China and shows potential for field applications in rock engineering.

HANDWRITTEN AMHARIC CHARACTER RECOGNITION SYSTEM USING CONVOLUTIONAL NEURAL NETWORKS

Amharic language is an official language of the federal government of the Federal Democratic Republic of Ethiopia. Accordingly, there is a bulk of handwritten Amharic documents available in libraries, information centres, museums, and offices. Digitization of these documents enables to harness already available language technologies to local information needs and developments. Converting these documents will have a lot of advantages including (i) to preserve and transfer history of the country (ii) to save storage space (ii) proper handling of documents (iv) enhance retrieval of information through internet and other applications. Handwritten Amharic character recognition system becomes a challenging task due to inconsistency of a writer, variability in writing styles of different writers, relatively large number of characters of the script, high interclass similarity, structural complexity and degradation of documents due to different reasons. In order to recognize handwritten Amharic character a novel method based on deep neural networks is used which has recently shown exceptional performance in various pattern recognition and machine learning applications, but has not been endeavoured for Ethiopic script. The Convolutional neural network model is evaluated for its performance using our database that contains 132,500 datasets of handwritten Amharic characters. Common handwritten recognition systems using machine learning use a combination of both feature extractors and classifiers. Currently the use of deep learning techniques shows promising improvements for machine learning based classification tasks. Our proposed CNN model gives an accuracy of 91.83% on training data and 90.47% on validation data.

An Efficient Convolutional Neural Network for Coronary Heart Disease Prediction

This study proposes an efficient neural network with convolutional layers to classify significantly class-imbalanced clinical data. The data are curated from the National Health and Nutritional Examination Survey (NHANES) with the goal of predicting the occurrence of Coronary Heart Disease (CHD). While the majority of the existing machine learning models that have been used on this class of data are vulnerable to class imbalance even after the adjustment of class-specific weights, our simple two-layer CNN exhibits resilience to the imbalance with fair harmony in class-specific performance. In order to obtain significant improvement in classification accuracy under supervised learning settings, it is a common practice to train a neural network architecture with a massive data and thereafter, test the resulting network on a comparatively smaller amount of data. However, given a highly imbalanced dataset, it is often challenging to achieve a high class 1 (true CHD prediction rate) accuracy as the testing data size increases. We adopt a two-step approach: first, we employ least absolute shrinkage and selection operator (LASSO) based feature weight assessment followed by majority-voting based identification of important features. Next, the important features are homogenized by using a fully connected layer, a crucial step before passing the output of the layer to successive convolutional stages. We also propose a training routine per epoch, akin to a simulated annealing process, to boost the classification accuracy. Despite a 35:1 (Non-CHD:CHD) ratio in the NHANES dataset, the investigation confirms that our proposed CNN architecture has the classification power of 77% to correctly classify the presence of CHD and 81.8% the absence of CHD cases on a testing data, which is 85.70% of the total dataset. This result signifies that the proposed architecture can be generalized to other studies in healthcare with a similar order of features and imbalances. While the recall values obtained from other machine learning methods such as SVM and random forest are comparable to that of our proposed CNN model, our model predicts the negative (Non-CHD) cases with higher accuracy. The balanced accuracy of our model (79.5%) is also better than individual accuracies of SVM or random forest classifiers. The CNN classifier results in high specificity and test accuracy along with high values of recall and area under the curve (AUC).

An Intrusion Detection Model Based on Feature Reduction and Convolutional Neural Networks

With the popularity and development of network technology and the Internet, intrusion detection systems (IDSs), which can identify attacks, have been developed. Traditional intrusion detection algorithms typically employ mining association rules to identify intrusion behaviors. However, they fail to fully extract the characteristic information of user behaviors and encounter various problems, such as high false alarm rate (FAR), poor generalization capability, and poor timeliness. In this paper, we propose a network intrusion detection model based on a convolutional neural network-IDS (CNN-IDS). Redundant and irrelevant features in the network traffic data are first removed using different dimensionality reduction methods. Features of the dimensionality reduction data are automatically extracted using the CNN, and more effective information for identifying intrusion is extracted by supervised learning. To reduce the computational cost, we convert the original traffic vector format into an image format and use a standard KDD-CUP99 dataset to evaluate the performance of the proposed CNN model. The experimental results indicate that the AC, FAR, and timeliness of the CNN-IDS model are higher than those of traditional algorithms. Therefore, the model we propose has not only research significance but also practical value.

Shoe-Print Image Retrieval With Multi-Part Weighted CNN

Identifying shoe-print impressions in the scene of crime (SoC) from database images is a challenging problem in forensic science due to the complicated impressing surface, the partial absence of on-site impressions, and the huge domain gap between the query and the gallery images. The existing approaches pay much attention to feature extraction while ignoring its distinctive characteristics. In this paper, we propose a novel multi-part weighted convolutional neural network (MP-CNN) for shoe-print image retrieval. Specifically, the proposed CNN model processes images in three steps: 1) dividing the input images vertically into two parts and extracting sub-features by a parameter-shared network individually; 2) calculating the importance weight matrix of the sub-features based on the informative pixels they contained and concatenating them as the final feature, and; 3) using the triplet loss function to measure the similarity between the query and the gallery images. In addition to the proposed network, we adopt an effective strategy to enhance the quality of the images and to reduce the domain gap using the U-Net structure. The experimental evaluations demonstrate that our proposed method significantly outperforms other fine-grained cross-domain methods on SPID dataset and obtains comparative results with the state-of-the-art shoe-print retrieval methods on FID300 dataset.

RESEARCH ON HIGH ACCURACY DETECTION OF RED TIDE HYPERSPECRRAL BASED ON DEEP LEARNING CNN

Abstract. Increasing frequency in red tide outbreaks has been reported around the world. It is of great concern due to not only their adverse effects on human health and marine organisms, but also their impacts on the economy of the affected areas. this paper put forward a high accuracy detection method based on a fully-connected deep CNN detection model with 8-layers to monitor red tide in hyperspectral remote sensing images, then make a discussion of the glint suppression method for improving the accuracy of red tide detection. The results show that the proposed CNN hyperspectral detection model can detect red tide accurately and effectively. The red tide detection accuracy of the proposed CNN model based on original image and filter-image is 95.58 % and 97.45 %, respectively, and compared with the SVM method, the CNN detection accuracy is increased by 7.52 % and 2.25 %. Compared with SVM method base on original image, the red tide CNN detection accuracy based on filter-image increased by 8.62 % and 6.37 %. It also indicates that the image glint affects the accuracy of red tide detection seriously.

A Neural-Based Stereoscopic Image Quality Assessment with Reference

Image quality is an important aspect for several applications (biometrics, tracking, object detection and so on). Several methods have been proposed in the literature to estimate it. These methods are able to predict subjective judgments according to different characteristics. The goal of this paper is to present a framework for stereoscopic image quality metric with reference based on Neural Networks (CNN & ANN). The proposed CNN model is composed of 3 convolutional layers and two Fully Connected (FC) layers and it is used to identify the degradation type in the image. The quality is then estimated using an ANN model. Its inputs are some computed features, selected according to the identified degradation type. The results obtained through two common datasets show the relevance of the proposed approach.