Deep Neural Network Classifier Research Articles

A novel skin cancer detection model using modified finch deep CNN classifier model

Skin cancer is one of the most life-threatening diseases caused by the abnormal growth of the skin cells, when exposed to ultraviolet radiation. Early detection seems to be more crucial for reducing aberrant cell proliferation because the mortality rate is rapidly rising. Although multiple researches are available based on the skin cancer detection, there still exists challenges in improving the accuracy, reducing the computational time and so on. In this research, a novel skin cancer detection is performed using a modified falcon finch deep Convolutional neural network classifier (Modified Falcon finch deep CNN) that efficiently detects the disease with higher efficiency. The usage of modified falcon finch deep CNN classifier effectively analyzed the information relevant to the skin cancer and the errors are also minimized. The inclusion of the falcon finch optimization in the deep CNN classifier is necessary for efficient parameter tuning. This tuning enhanced the robustness and boosted the convergence of the classifier that detects the skin cancer in less stipulated time. The modified falcon finch deep CNN classifier achieved accuracy, sensitivity, and specificity values of 93.59%, 92.14%, and 95.22% regarding k-fold and 96.52%, 96.69%, and 96.54% regarding training percentage, proving more effective than literary works.

Trusted Perception Method for Traffic Signs That Are Physically Attacked

Traffic sign recognition is a crucial method by which autonomous driving systems acquire road information, and is predominantly based on deep neural networks (DNNs). However, the recognition results of DNNs are not always trustworthy for traffic signs subject to abnormal disturbance. Recently, the phenomenon of adversarial examples successfully deceiving DNNs has garnered considerable attention. Because DNN-based computer vision techniques are becoming increasingly prevalent in traffic scenarios, the misclassification of attacked traffic signs by DNN classifiers poses serious safety hazards. Although numerous methods have been proposed for crafting physical adversarial examples that are robust in the real world, most existing defense approaches focus on digital attacks, which necessitate the adversary infiltrating the embedded system; thus, it becomes challenging to obtain results. A reliable approach for defending against physical adversarial traffic signs enables autonomous vehicles to achieve trusted perception of traffic signs. In this paper, we present a deep image prior-based pipeline to defend against robust adversarial traffic signs in the real world, an approach that circumvents the need for prior data sets during training. Our approach protects the safety of autonomous vehicles by performing image reconstruction of captured traffic sign images. The genuine traffic sign class can be inferred by leveraging the consistency of the victim classifier’s decision results for reconstructed images at different stages. Additionally, we evaluate the efficacy of our defense pipeline for detecting other potential types of physical adversarial traffic signs that may exist in the real world, thus demonstrating the generalizability of our approach.

An Automatic Method to Prevent and Classify Cyberbullying Incidents using Machine Learning Approach

The technological advancements and the increasing popularity of social networking platforms, the sharing of personal information among online users has become widespread. This sharing occurs effortlessly through various devices such as computers and mobile phones. Cyberbullying can manifest through SMS, text messages, and various applications, as well as online platforms like social media and forums, where individuals can view, engage with, or distribute content.The project offers a comprehensive understanding of Cyberbullying incidents and their corresponding offences combining a series of approaches reported in relevant Work. The implementation provides the opportunity to systematically combine various element or Cyberbullying characteristics. Additionally, a comprehensive list of Cyberbullying-related offences is put forward. The offenses are ordered in a Deep Neural Network classification system based on specific criteria to assist in better classification and correlation of their respective incidents. This enables a thorough understanding of the repeating and underlying criminal activities. This study focuses on classifying user posts and image content into bullying or non bullying through reputation score

Joint negative–positive-learning based sample reweighting for hyperspectral image classification with label noise

Deep neural networks (DNNs) have been widely used for hyperspectral image (HSI) classification. However, the superior performance of DNNs requires accurately labeled samples. In various applications, label noise in large-scale HSI data is often unavoidable and adversely affects the performance of a classifier. Therefore, establishing a robust classification model from an existing DNN classifier for the HSI dataset with label noise is a significant and challenging problem. In this report, we propose a meta-learning reweighting framework based on a joint positive and negative learning (JPNL) method which can adaptively reweight labeled samples to improve the robustness of the classification model. Experimental results on two commonly used hyperspectral imaging (HSI) datasets contaminated with label noise demonstrate significant improvements in classification performance and robustness achieved by the proposed framework, as compared to existing classifiers.

Data leakage in deep learning studies of translational EEG.

A growing number of studies apply deep neural networks (DNNs) to recordings of human electroencephalography (EEG) to identify a range of disorders. In many studies, EEG recordings are split into segments, and each segment is randomly assigned to the training or test set. As a consequence, data from individual subjects appears in both the training and the test set. Could high test-set accuracy reflect data leakage from subject-specific patterns in the data, rather than patterns that identify a disease? We address this question by testing the performance of DNN classifiers using segment-based holdout (in which segments from one subject can appear in both the training and test set), and comparing this to their performance using subject-based holdout (where all segments from one subject appear exclusively in either the training set or the test set). In two datasets (one classifying Alzheimer's disease, and the other classifying epileptic seizures), we find that performance on previously-unseen subjects is strongly overestimated when models are trained using segment-based holdout. Finally, we survey the literature and find that the majority of translational DNN-EEG studies use segment-based holdout. Most published DNN-EEG studies may dramatically overestimate their classification performance on new subjects.

Speech emotion recognition using a combination of variational mode decomposition and Hilbert transform

The objective of automated speech emotion recognition (SER) is to recognize each emotion of the speech signal uniquely and efficiently using machines like computers, mobile devices, etc. The popularity of SER has been widely acknowledged among researchers due to its extensive applicability in practical contexts. The use of SER has shown to be advantageous in several domains, including medical treatment, enhancement of security systems, surveillance operations, online marketing strategies, online educational platforms, online search engines, personal communication, customer relationship management, reinforcement of machine and human connection, and numerous other areas. Numerous authors have used various techniques, including the combination of multiple features (acoustic, non-acoustic, or both) and classifiers (machine learning, deep learning, or both), in order to enhance the efficacy of emotion categorization. In the present work, our objective is to enhance the performance of emotion classification (PECL) by using a combined approach of the variational mode decomposition (VMD) and Hilbert transform (HT) techniques. Using the VMD method, we decomposed the speech signal frame into many sub-signals, or intrinsic mode functions (IMFs). Then HT is applied to each VMD-based IMF signal to find the mode instantaneous amplitude (MIA) and mode instantaneous frequency (MIF) signal vectors. We extracted proposed features such as HT-based approximate entropy (HTAE), HT-based permutation entropy (HTPE), HT-based increment entropy (HTIE), and HT-based sample entropy (HTSE) using each MIA and MIF signal vector. The combination of the proposed HT-based features is called HT-based entropy (HTE) features. Then, we accessed the PECL using the HTE and MFCC features alone and in conjunction with a deep neural network (DNN) classifier. The experiment's results showed that the combinations of the proposed feature (MFCC + HTE) using a DNN classifier outperformed the individual features and obtained a SER accuracy of 86.92% for the EMOVO dataset and 91.63% for the EMO-DB dataset.

Learning with noisy labels using collaborative sample selection and contrastive semi-supervised learning

Learning with noisy labels (LNL) has been extensively studied, with existing approaches typically following a framework that alternates between clean sample selection and semi-supervised learning (SSL). However, this approach has a limitation: the clean set selected by the Deep Neural Network (DNN) classifier, trained through self-training, inevitably contains noisy samples. This mixture of clean and noisy samples leads to misguidance in DNN training during SSL, resulting in impaired generalization performance due to confirmation bias caused by error accumulation in sample selection. To address this issue, we propose a method called Collaborative Sample Selection (CSS), which leverages the large-scale pre-trained model CLIP. CSS aims to remove the mixed noisy samples from the identified clean set. We achieve this by training a 2-Dimensional Gaussian Mixture Model (2D-GMM) that combines the probabilities from CLIP with the predictions from the DNN classifier. To further enhance the adaptation of CLIP to LNL, we introduce a co-training mechanism with a contrastive loss in semi-supervised learning. This allows us to jointly train the prompt of CLIP and the DNN classifier, resulting in improved feature representation, boosted classification performance of DNNs, and reciprocal benefits to our Collaborative Sample Selection. By incorporating auxiliary information from CLIP and utilizing prompt fine-tuning, we effectively eliminate noisy samples from the clean set and mitigate confirmation bias during training. Experimental results on multiple benchmark datasets demonstrate the effectiveness of our proposed method in comparison with the state-of-the-art approaches.

A Hybrid Approach for Binary Classification of Imbalanced Data

Binary classification with an imbalanced dataset is challenging. Models tend to consider all samples as belonging to the majority class. Although existing solutions such as sampling methods, cost-sensitive methods, and ensemble learning methods improve the poor accuracy of the minority class, these methods are limited by overfitting or cost parameters that are difficult to decide. This paper proposes a hybrid approach with dimension reduction that consists of data block construction, dimensionality reduction, and ensemble learning with deep neural network classifiers. The performance is evaluated on eight imbalanced public datasets in terms of recall, G-mean, AUC, F-measure, and balanced accuracy. The results show that the proposed model outperforms state-of-the-art methods.

Efficient Target Classification Based on Vehicle Volume Estimation in High-Resolution Radar Systems

In this paper, we propose a method for efficient target classification based on the spatial features of the point cloud generated by using a high-resolution radar sensor. The frequency-modulated continuous wave radar sensor can estimate the distance and velocity of a target. In addition, the azimuth and elevation angle of the target can be estimated by using a multiple-input and multiple-output antenna system. Using the estimated distance, velocity, and angle, the 3D point cloud of target can be generated. From the generated point cloud, we extract the point cloud for each individual target using the density-based spatial clustering of application with noise method and a camera mounted on the radar sensor. Then, we define the convex hull boundaries that enclose these point clouds in both 3D and 2D spaces obtained by orthogonally projecting onto the xy, yz, and zx planes. Using the vertices of convex hull, we calculate the volume of the targets and the areas in 2D spaces. Several feature points, including the calculated spatial information, are numerized and configured into feature vectors. We design an uncomplicated deep neural network classifier based on minimal input information to achieve fast and efficient classification performance. As a result, the proposed method achieved an average accuracy of 97.1%, and the time required for training was reduced compared to the method using only point cloud data and the convolutional neural network-based method.

Artificial intelligence-enhanced automation for M-mode echocardiographic analysis: ensuring fully automated, reliable, and reproducible measurements.

To enhance M-mode echocardiography's utility for measuring cardiac structures, we developed and evaluated an artificial intelligence (AI)-based automated analysis system for M-mode images through the aorta and left atrium [M-mode (Ao-LA)], and through the left ventricle [M-mode (LV)]. Our system, integrating two deep neural networks (DNN) for view classification and image segmentation, alongside an auto-measurement algorithm, was developed using 5,958M-mode images [3,258M-mode (LA-Ao), and 2,700M-mode (LV)] drawn from a nationwide echocardiographic dataset collated from five tertiary hospitals. The performance of view classification and segmentation DNNs were evaluated on 594M-mode images, while automatic measurement accuracy was tested on separate internal test set with 100M-mode images as well as external test set with 280 images (140 sinus rhythm and 140 atrial fibrillation). Performance evaluation showed the view classification DNN's overall accuracy of 99.8% and segmentation DNN's Dice similarity coefficient of 94.3%. Within the internal test set, all automated measurements, including LA, Ao, and LV wall and cavity, resonated strongly with expert evaluations, exhibiting Pearson's correlation coefficients (PCCs) of 0.81-0.99. This performance persisted in the external test set for both sinus rhythm (PCC, 0.84-0.98) and atrial fibrillation (PCC, 0.70-0.97). Notably, automatic measurements, consistently offering multi-cardiac cycle readings, showcased a stronger correlation with the averaged multi-cycle manual measurements than with those of a single representative cycle. Our AI-based system for automatic M-mode echocardiographic analysis demonstrated excellent accuracy, reproducibility, and speed. This automated approach has the potential to improve efficiency and reduce variability in clinical practice.

Multiclass classification of diseased grape leaf identification using deep convolutional neural network(DCNN) classifier

The cultivation of grapes encounters various challenges, such as the presence of pests and diseases, which have the potential to considerably diminish agricultural productivity. Plant diseases pose a significant impediment, resulting in diminished agricultural productivity and economic setbacks, thereby affecting the quality of crop yields. Hence, the precise and timely identification of plant diseases holds significant importance. This study employs a Convolutional neural network (CNN) with and without data augmentation, in addition to a DCNN Classifier model based on VGG16, to classify grape leaf diseases. A publicly available dataset is utilized for the purpose of investigating diseases affecting grape leaves. The DCNN Classifier Model successfully utilizes the strengths of the VGG16 model and modifies it by incorporating supplementary layers to enhance its performance and ability to generalize. Systematic evaluation of metrics, such as accuracy and F1-score, is performed. With training and test accuracy rates of 99.18 and 99.06%, respectively, the DCNN Classifier model does a better job than the CNN models used in this investigation. The findings demonstrate that the DCNN Classifier model, utilizing the VGG16 architecture and incorporating three supplementary CNN layers, exhibits superior performance. Also, the fact that the DCNN Classifier model works well as a decision support system for farmers is shown by the fact that it can quickly and accurately identify grape diseases, making it easier to take steps to stop them. The results of this study provide support for the reliability of the DCNN classifier model and its potential utility in the field of agriculture.

DDA-SSNets: Dual decoder attention-based semantic segmentation networks for COVID-19 infection segmentation and classification using chest X-Ray images.

COVID-19 needs to be diagnosed and staged to be treated accurately. However, prior studies' diagnostic and staging abilities for COVID-19 infection needed to be improved. Therefore, new deep learning-based approaches are required to aid radiologists in detecting and quantifying COVID-19-related lung infections. To develop deep learning-based models to classify and quantify COVID-19-related lung infections. Initially, Dual Decoder Attention-based Semantic Segmentation Networks (DDA-SSNets) such as Dual Decoder Attention-UNet (DDA-UNet) and Dual Decoder Attention-SegNet (DDA-SegNet) are proposed to facilitate the dual segmentation tasks such as lung lobes and infection segmentation in chest X-ray (CXR) images. The lung lobe and infection segmentations are mapped to grade the severity of COVID-19 infection in both the lungs of CXRs. Later, a Genetic algorithm-based Deep Convolutional Neural Network classifier with the optimum number of layers, namely GADCNet, is proposed to classify the extracted regions of interest (ROI) from the CXR lung lobes into COVID-19 and non-COVID-19. The DDA-SegNet shows better segmentation with an average BCSSDC of 99.53% and 99.97% for lung lobes and infection segmentations, respectively, compared with DDA-UNet with an average BCSSDC of 99.14% and 99.92%. The proposed DDA-SegNet with GADCNet classifier offered excellent classification results with an average BCCAC of 99.98%, followed by the GADCNet with DDA-UNet with an average BCCAC of 99.92% after extensive testing and analysis. The results show that the proposed DDA-SegNet has superior performance in the segmentation of lung lobes and COVID-19-infected regions in CXRs, along with improved severity grading compared to the DDA-UNet and improved accuracy of the GADCNet classifier in classifying the CXRs into COVID-19, and non-COVID-19.

Early detection of infestation by mustard aphid, vegetable thrips and two-spotted spider mite in bok choy with deep neural network (DNN) classification model using hyperspectral imaging data

Although exclusion measures (e.g., air filters, biosecurity practices) can be employed to prevent occurrence of pest outbreaks, indoors vegetable farms in Singapore are still susceptible to various arthropod pests. Due to strong interest from the industry to pursue pesticide-free production, indoors pest management is often focused on early detection for timely containment and eradication, implying the importance of robust and vigorous pest monitoring programs. In recent years, application of machine vision technologies, especially hyperspectral imaging (HSI), has been studied for their capacity to early detect pest infestation. However, there is a lack of studies conducted in actual indoor environments and on multiple arthropod pests. Thus, this study aimed to non-destructively collect hyperspectral data of bok choy Brassica rapa subspecies chinensis which were healthy or infested with either mustard aphids Lipaphis erisymi, vegetable thrips Echinothrips americanus or two-spotted spider mites Tetranichus urticae in indoor environment to build deep neural network (DNN) classification model for early detection. Based on HSI data of control and infested plants collected daily over a period of two weeks, we found that point percentage change (PPC) values associated with leaf reflectance in 420–440 nm, 500–520 nm, 620–637 nm, 720–800 nm, and 850 nm were sensitive to infestation by the mentioned arthropod pests. Deep Neural Network (DNN) classification models trained on collected HSI data were found to outperform Linear Discriminant Analysis (LDA) and Support Vector Machine (SVM) classification models. DNN models achieved 92.8 ± 0.4 % overall classification accuracy across all days. As early as two days after infestation, DNN models could achieved classification precision values of 96.4 %, 96.9 %, 93.9 % and 100 % for control plants and plants infested with either aphids, spider mites or thrips respectively. These results highlight the feasibility of multiclass early detection of different arthropod pests and the potential of HSI system coupled with DNN classification as an autonomous plant health monitoring tool in indoor crop production.

An Integrated Machine Learning Approach for Congestive Heart Failure Prediction.

Congestive heart failure (CHF) is one of the primary sources of mortality and morbidity among the global population. Over 26 million individuals globally are affected by heart disease, and its prevalence is rising by 2% yearly. With advances in healthcare technologies, if we predict CHF in the early stages, one of the leading global mortality factors can be reduced. Therefore, the main objective of this study is to use machine learning applications to enhance the diagnosis of CHF and to reduce the cost of diagnosis by employing minimum features to forecast the possibility of a CHF occurring. We employ a deep neural network (DNN) classifier for CHF classification and compare the performance of DNN with various machine learning classifiers. In this research, we use a very challenging dataset, called the Cardiovascular Health Study (CHS) dataset, and a unique pre-processing technique by integrating C4.5 and K-nearest neighbor (KNN). While the C4.5 technique is used to find significant features and remove the outlier data from the dataset, the KNN algorithm is employed for missing data imputation. For classification, we compare six state-of-the-art machine learning (ML) algorithms (KNN, logistic regression (LR), naive Bayes (NB), random forest (RF), support vector machine (SVM), and decision tree (DT)) with DNN. To evaluate the performance, we use seven statistical measurements (i.e., accuracy, specificity, sensitivity, F1-score, precision, Matthew's correlation coefficient, and false positive rate). Overall, our results reflect our proposed integrated approach, which outperformed other machine learning algorithms in terms of CHF prediction, reducing patient expenses by reducing the number of medical tests. The proposed model obtained 97.03% F1-score, 95.30% accuracy, 96.49% sensitivity, and 97.58% precision.

Intelligent vegetable freshness monitoring system developed by integrating eco-friendly fluorescent sensor arrays with deep convolutional neural networks

Given the perishable and seasonal nature of vegetables, monitoring their freshness is essential to ensure food safety and reduce waste. Currently, there are limited packaging systems for fresh vegetables that incorporate intelligent freshness monitoring labels. Herein, we report on the development and application of a 3 × 6 fluorescent sensor array that exhibits pH-sensitive properties, utilizing curcumin, puerarin, and fisetin. During spoilage, yardlong beans and spinach, which had high protein content, produced alkaline volatile organic compounds (VOCs), whereas sweet corn, rich in sugar, emitted acidic VOCs. The fluorescent sensor array, integrated with deep convolutional neural network (DCNN), enabled non-destructive, real-time, and accurate classification of the freshness of the aforementioned three vegetables by detecting the acidity or alkalinity of their VOCs. The trained ResNet50 DCNN model achieved an overall accuracy of 96.21 % in classifying the freshness of the aforementioned vegetables in the testing set, with specific accuracies of 98.58 % for yardlong beans, 97.15 % for spinach, and 92.89 % for sweet corn, respectively. This intelligent freshness monitoring platform is adaptable for monitoring and classifying the freshness of a wide range of agricultural and food products.

Using Longitudinal Twitter Data for Digital Epidemiology of Childhood Health Outcomes: An Annotated Data Set and Deep Neural Network Classifiers.

We manually annotated 9734 tweets that were posted by users who reported their pregnancy on Twitter, and used them to train, evaluate, and deploy deep neural network classifiers (F1-score=0.93) to detect tweets that report having a child with attention-deficit/hyperactivity disorder (678 users), autism spectrum disorders (1744 users), delayed speech (902 users), or asthma (1255 users), demonstrating the potential of Twitter as a complementary resource for assessing associations between pregnancy exposures and childhood health outcomes on a large scale.

Towards Model Extraction Attacks in GAN-Based Image Translation via Domain Shift Mitigation

Model extraction attacks (MEAs) enable an attacker to replicate the functionality of a victim deep neural network (DNN) model by only querying its API service remotely, posing a severe threat to the security and integrity of pay-per-query DNN-based services. Although the majority of current research on MEAs has primarily concentrated on neural classifiers, there is a growing prevalence of image-to-image translation (I2IT) tasks in our everyday activities. However, techniques developed for MEA of DNN classifiers cannot be directly transferred to the case of I2IT, rendering the vulnerability of I2IT models to MEA attacks often underestimated. This paper unveils the threat of MEA in I2IT tasks from a new perspective. Diverging from the traditional approach of bridging the distribution gap between attacker queries and victim training samples, we opt to mitigate the effect caused by the different distributions, known as the domain shift. This is achieved by introducing a new regularization term that penalizes high-frequency noise, and seeking a flatter minimum to avoid overfitting to the shifted distribution. Extensive experiments on different image translation tasks, including image super-resolution and style transfer, are performed on different backbone victim models, and the new design consistently outperforms the baseline by a large margin across all metrics. A few real-life I2IT APIs are also verified to be extremely vulnerable to our attack, emphasizing the need for enhanced defenses and potentially revised API publishing policies.

Enhancing cancer stage prediction through hybrid deep neural networks: a comparative study.

Efficiently detecting and treating cancer at an early stage is crucial to improve the overall treatment process and mitigate the risk of disease progression. In the realm of research, the utilization of artificial intelligence technologies holds significant promise for enhancing advanced cancer diagnosis. Nonetheless, a notable hurdle arises when striving for precise cancer-stage diagnoses through the analysis of gene sets. Issues such as limited sample volumes, data dispersion, overfitting, and the use of linear classifiers with simple parameters hinder prediction performance. This study introduces an innovative approach for predicting early and late-stage cancers by integrating hybrid deep neural networks. A deep neural network classifier, developed using the open-source TensorFlow library and Keras network, incorporates a novel method that combines genetic algorithms, Extreme Learning Machines (ELM), and Deep Belief Networks (DBN). Specifically, two evolutionary techniques, DBN-ELM-BP and DBN-ELM-ELM, are proposed and evaluated using data from The Cancer Genome Atlas (TCGA), encompassing mRNA expression, miRNA levels, DNA methylation, and clinical information. The models demonstrate outstanding prediction accuracy (89.35%-98.75%) in distinguishing between early- and late-stage cancers. Comparative analysis against existing methods in the literature using the same cancer dataset reveals the superiority of the proposed hybrid method, highlighting its enhanced accuracy in cancer stage prediction.

Artifact removal and motor imagery classification in EEG using advanced algorithms and modified DNN

This paper presents an advanced approach for EEG artifact removal and motor imagery classification using a combination of Four Class Iterative Filtering and Filter Bank Common Spatial Pattern Algorithm with a Modified Deep Neural Network (DNN) classifier. The research aims to enhance the accuracy and reliability of BCI systems by addressing the challenges posed by EEG artifacts and complex motor imagery tasks.The methodology begins by introducing FCIF, a novel technique for ocular artifact removal, utilizing iterative filtering and filter banks. FCIF's mathematical formulation allows for effective artifact mitigation, thereby improving the quality of EEG data. In tandem, the FC-FBCSP algorithm is introduced, extending the Filter Bank Common Spatial Pattern approach to handle four-class motor imagery classification. The Modified DNN classifier enhances the discriminatory power of the FC-FBCSP features, optimizing the classification process.The paper showcases a comprehensive experimental setup, featuring the utilization of BCI Competition IV Dataset 2a & 2b. Detailed preprocessing steps, including filtering and feature extraction, are presented with mathematical rigor. Results demonstrate the remarkable artifact removal capabilities of FCIF and the classification prowess of FC-FBCSP combined with the Modified DNN classifier. Comparative analysis highlights the superiority of the proposed approach over baseline methods and the method achieves the mean accuracy of 98.575%.

A Novel Assessment of Lung Cancer Classification System Using Binary Grasshopper with Artificial Bee Optimisation Algorithm with Double Deep Neural Network Classifier

A Novel Assessment of Lung Cancer Classification System Using Binary Grasshopper with Artificial Bee Optimisation Algorithm with Double Deep Neural Network Classifier