CNN-based Feature Extraction Research Articles

Body gesture recognition using crow search algorithm enhanced probabilistic neural network for human- computer interaction

Background Body gesture recognition has become a fundamental technique in Human-Computer Interaction (HCI). As human-machine interaction evolves, there is an increasing need for precise and efficient gesture detection systems. However, current methods face limitations such as accuracy constraints, high computational complexity, and limited adaptability. This study addresses these challenges by proposing an innovative approach to enhance the accuracy and efficiency of body gesture recognition systems. Methods The proposed system integrates advanced algorithms and techniques to improve performance. A Marker-Based Watershed Algorithm is employed for accurate image segmentation, enhancing region detection. Feature extraction uses a Convolutional Neural Network (CNN), while a Wavelet Transform-Based Pre-Processing technique improves input data quality. A unique component of this method is the application of the Crow Search Algorithm to optimize model efficiency. An Optimized Probabilistic Neural Network (PNN) is utilized for gesture classification, aiming to increase precision and computational effectiveness. Results The proposed approach achieves a gesture recognition accuracy rate of 99%. Compared to traditional methods such as Decision Trees (DT), Support Vector Machines (SVM), and Improved Neural Networks (INN), the Optimized PNN demonstrates a 2.21% improvement in overall accuracy. The implementation, carried out in Python, showcases the robustness and adaptability of the system across diverse HCI applications. Conclusions This work presents a comprehensive solution to the challenges of body gesture recognition by integrating cutting-edge algorithms. Combining the Marker-Based Watershed Algorithm, CNN-based feature extraction, and Crow Search Optimization significantly enhances the system’s accuracy and efficiency. By addressing the shortcomings of existing methods, this approach provides a more responsive, reliable, and flexible gesture recognition system, contributing to the advancement of HCI technologies. The results demonstrate the potential for improved human-computer interaction through more effective and precise gesture detection.

Comparative Analysis of CNN-based Feature Extraction Techniques and Conventional Machine Learning Models for Quality Assessment in Agricultural Produce

Comparative Analysis of CNN-based Feature Extraction Techniques and Conventional Machine Learning Models for Quality Assessment in Agricultural Produce

Convolutional Neural Network-based Smart Disaster Management Framework for Real-time Detection and Management of Forest Fires

Introduction: Timely detection of catastrophic natural disasters, such as forest fires, is critical to minimizing losses and ensuring rapid response. Artificial intelligence is increasingly being recognized as a valuable tool in enhancing various stages of disaster management. Method: This paper presents the development of a smart framework utilizing machine learning techniques for real-time detection and monitoring of natural disasters, specifically forest fires. The proposed approach employs a 10-layer convolutional neural network (CNN) that classifies aerial images into Fire, Non-Fire, and Smoke categories with high precision and speed. In addition to this, a CNN-based feature extraction process is performed and integrated with various ML classifiers, including support vector machine, k-nearest neighbor, decision tree, random forest, and extra trees. Results: Extensive performance analysis reveals that the proposed 10-layer CNN model outperforms other classifiers, achieving an accuracy of 97.64% in the binary classification of fire vs. nonfire and 95.61% in the three-class classification of Fire, Non-Fire and Smoke classes. Furthermore, a comparative study with existing state-of-the-art methods demonstrates the proposed model's superior performance in both accuracy and computational complexity. Conclusion: These results demonstrate the potential of the proposed CNN-based framework to serve as a reliable and effective tool for real-time disaster management across various applications, providing valuable support to emergency response teams in mitigating the impact of natural disasters.

Enhancing Movie Recommendation Systems Through CNN-Based Feature Extraction and Optimized Collaborative Filtering

Abstract. This paper presents the design and implementation of a movie recommendation model that combines collaborative filtering with Convolutional Neural Networks (CNNs) to tackle the challenge of information overload in extensive movie databases. The primary objective is to enhance the user experience by developing an efficient and personalized recommendation system. The proposed model integrates CNNs to extract detailed image features from movie posters, enriching the movie representations used in the recommendation process. This feature extraction is coupled with an optimized collaborative filtering algorithm to deliver more accurate and tailored movie suggestions. The study is based on a dataset that includes movie information, user ratings, and user behavior data. Experimental results demonstrate that the proposed model significantly improves the accuracy and personalization of movie recommendations, thereby offering users a superior selection experience. The practical implications of this research are substantial, as it has the potential to boost user satisfaction and engagement on movie recommendation platforms. Future work will focus on further enhancing the model by incorporating long-term user behavior patterns and integrating multi-modal data sources for even more precise recommendations.

Analysis of Lung Disease Prediction using Machine Learning Algorithms

Lung diseases are a notable in global health concern, requiring early diagnosis for better recovery and survival rates. Deep learning strategies, especially CNNs, have shown great promise in self learning lung disease diagnosis from medical images like chest X-rays. Ensemble learning methods using pretrained networks such as VGG16, InceptionV3, and MobileNetV2 have achieved up to 94% accuracy in identifying conditions like COVID-19, pneumonia, and lung opacity. Lightweight CNN models also performed well, with accuracy up to 89.89%. Traditional machine learning algorithms, including Random Forest and Logistic Regression, yielded accuracy rates between 88% and 90%. A hybrid deep learning approach, combining CNN based feature extraction with classifiers like AdaBoost, SVM, and Random Forest, improved classification accuracy by 3.1% and reduced computational complexity by 16.91%. This hybrid method highlights the main feature for integrating deep learning with traditional classifiers to enhance lung disease detection efficiency

DNASimCLR: a contrastive learning-based deep learning approach for gene sequence data classification

BackgroundThe rapid advancements in deep neural network models have significantly enhanced the ability to extract features from microbial sequence data, which is critical for addressing biological challenges. However, the scarcity and complexity of labeled microbial data pose substantial difficulties for supervised learning approaches. To address these issues, we propose DNASimCLR, an unsupervised framework designed for efficient gene sequence data feature extraction.ResultsDNASimCLR leverages convolutional neural networks and the SimCLR framework, based on contrastive learning, to extract intricate features from diverse microbial gene sequences. Pre-training was conducted on two classic large scale unlabelled datasets encompassing metagenomes and viral gene sequences. Subsequent classification tasks were performed by fine-tuning the pretrained model using the previously acquired model. Our experiments demonstrate that DNASimCLR is at least comparable to state-of-the-art techniques for gene sequence classification. For convolutional neural network-based approaches, DNASimCLR surpasses the latest existing methods, clearly establishing its superiority over the state-of-the-art CNN-based feature extraction techniques. Furthermore, the model exhibits superior performance across diverse tasks in analyzing biological sequence data, showcasing its robust adaptability.ConclusionsDNASimCLR represents a robust and database-agnostic solution for gene sequence classification. Its versatility allows it to perform well in scenarios involving novel or previously unseen gene sequences, making it a valuable tool for diverse applications in genomics.

A novel low-cost bearing fault diagnosis method based on convolutional neural network with full stage optimization in strong noise environment

In recent years, convolutional neural network (CNN) has been successfully applied in the field of bearing fault diagnosis. So as to improve the diagnosis performance in harsh environment with strong noise, the structure of CNN-based feature extractor becomes deeper and more complex. However, with the increase of depth, the model may lose shallow features and the training parameters will surge. Moreover, if the sample size is not large, it tends to over fit. It deviates from the concept of network lightweight. On the other hand, little attention will be paid to the optimization of model classifiers which can significantly improve the classification performance. Therefore, we proposed a CNN with full stage optimization (FSOCNN) model for bearing fault diagnosis in strong noise environment. In the feature extraction stage, the model is optimized with a novel multi-feature output structure connected with global average pooling to improve the feature extraction ability without any extra trainable parameters. In the classification stage, the traditional softmax layer will only participate in the parameter optimization of CNN model through gradient descent algorithm, and the diagnosis results will be output by support vector machine. The effectiveness of the proposed method is verified on the two bearing datasets under different levels of noise. Compared with the existing five fault diagnosis models, the results prove that the proposed method possesses higher accuracy, less computing time, and better stability.

Leveraging 3D convolutional neural network and 3D visible-near-infrared multimodal imaging for enhanced contactless oximetry.

Monitoring oxygen saturation ( ) is important in healthcare, especially for diagnosing and managing pulmonary diseases. Non-contact approaches broaden the potential applications of measurement by better hygiene, comfort, and capability for long-term monitoring. However, existing studies often encounter challenges such as lower signal-to-noise ratios and stringent environmental conditions. We aim to develop and validate a contactless measurement approach using 3D convolutional neural networks (3D CNN) and 3D visible-near-infrared (VIS-NIR) multimodal imaging, to offer a convenient, accurate, and robust alternative for monitoring. We propose an approach that utilizes a 3D VIS-NIR multimodal camera system to capture facial videos, in which is estimated through 3D CNN by simultaneously extracting spatial and temporal features. Our approach includes registration of multimodal images, tracking of the 3D region of interest, spatial and temporal preprocessing, and 3D CNN-based feature extraction and regression. In a breath-holding experiment involving 23 healthy participants, we obtained multimodal video data with reference values ranging from 80% to 99% measured by pulse oximeter on the fingertip. The approach achieved a mean absolute error (MAE) of 2.31% and a Pearson correlation coefficient of 0.64 in the experiment, demonstrating good agreement with traditional pulse oximetry. The discrepancy of estimated values was within 3% of the reference for of all 1-s time points. Besides, in clinical trials involving patients with sleep apnea syndrome, our approach demonstrated robust performance, with an MAE of less than 2% in estimations compared to gold-standard polysomnography. The proposed approach offers a promising alternative for non-contact oxygen saturation measurement with good sensitivity to desaturation, showing potential for applications in clinical settings.

Few-shot image classification via hybrid representation

Few-shot image classification aims to learn an embedding model on the base datasets and design a base learner to recognize novel categories. The few-shot image classification framework is a two-phase process. First, the pre-train phase utilizes the base data to train a CNN-based feature extractor. Next, in the meta-test phase, the frozen feature extractor is applied to novel data with categories different from the base data. A base learner is then designed for recognition. Several simple base learners, including nearest neighbor, support vector machine, and logistic regression classifiers, have been recently introduced for few-shot learning tasks. However, these base learners are separately designed to consider specific representations (e.g., the class center) or shared representations (e.g., the boundaries). This paper mainly focuses on exploring the representation-residual base learners, which aim to represent a query sample with the support set and predict the query sample’s label based on the minimal residual error. We first introduce two representation-residual base learners: a specific representation base learner and a shared representation base learner. Then, we propose a novel hybrid representation base learner that combines both base learners to generate competitive representation. Additionally, we extend our approach by incorporating a self-training framework to utilize the query data fully. We evaluate our proposed method on several benchmark few-shot image classification datasets, such as miniImageNet, tieredImageNet, CIFAR-FS, FC100, and CUB datasets. The experimental results indicate that our proposed approach shows a significant performance improvement.

Multi-dimensional CNN Based Feature Extraction with Feature Fusion and SVM for Human Activity Recognition in Surveillance Videos

Background/Objectives: The accurate recognition of human activities from video sequences is very challenging due to low resolution, cluttered background, partial occlusion, and different viewpoints. Machine learning (ML) based automated HAR from surveillance videos is required with the fusion of various feature extraction techniques. Methods: In this paper, SVM with feature fusion is utilized for automatic recognition from surveillance videos. A Histogram of Oriented Gradient (HOG) is used to segment the frame to differentiate humans from other objects or background noise in the input video frames. The multi-feature extraction can be accomplished in terms of Gabor Wavelet Transform (GWT), Autocorrelogram, Gray-Level Co-Occurrence Matrix (GLCM), HSV histogram, and Multi-dimensional CNN. The proposed approach is implemented in MATLAB software and compared with existing approaches like Space-Time Interest Point (STIP) and Histogram of Optical Flow (HOF). Findings: The proposed approach outperforms the existing approaches in terms of reduced time consumption and high accuracy, 99.886% when using the UCF101 dataset and 99.538% when using the UTKinect dataset. Novelty: The most discriminative feature information is obtained with the feature-level fusion technique. From the feature information, various human actions are recognized with the classification algorithm. Keywords: Human activity recognition, Machine Learning, Surveillance Videos, Human detection algorithm, Feature extraction, SVM classifier

Enhancing diabetic retinopathy detection through preprocessing and feature extraction with MGA-CSG algorithm

Anticipatory monitoring of diabetic retinal (DR) disease is crucial in preventing vision loss and blindness, making it a leading cause of worldwide vision impairment. In this study, we propose a novel technique, the Modified Generative Adversarial-based Crossover Salp Grasshopper (MGA-CSG) approach, for early prediction and accurate classification of diabetic retinal diseases using fundus image datasets. The proposed MGA-CSG approach combines deep learning with an optimization algorithm to achieve superior classification accuracy. We start with pre-processing of data to attenuate image variation, convert intensities, denoise, and enhance contrast in the fundus images. Next, a Convolutional Neural Network (CNN) is used for feature extraction, capturing essential information from the images. To address the limited amount of feature vectors generated by the CNN and improve learning uncertainty, we employ a Generative Adversarial Network (GAN) model to augment the feature vectors. It generates additional images, allowing the classifier to train with fewer input samples, leading to improved classification performance. One of the key novelties of our approach lies in the utilization of both labeled and unlabeled data samples, enhancing the GAN training performance. We replace the binary classifier in the GAN with a multiclass classifier, enabling separate discrimination of retinal disease classes. The MGA-CSG approach leverages the crossover Grasshopper Optimizer Algorithm (GOA) and Salp Swarm Algorithm (SSA) to optimize the model and avoid falling into local optimal solutions. This optimization process further enhances the accuracy of diabetic retinal disease classification. Experimental results demonstrate the superiority of the proposed MGA-CSG approach, achieving high accuracy, precision, recall, specificity, F1-measure, and Kappa coefficient in predicting glaucoma, diabetic retinopathy, cataract, macular edema, and myopia. The classification achieved impressive accuracy, precision, recall, specificity, F1-measure, and Kappa coefficient of 98.8%, 98.7%, 96.5%, 97.1%, 98.2%, and 95%, respectively. Overall, our novel MGA-CSG approach offers an efficient and accurate method for early screening and classification of diabetic retinal diseases. The combination of CNN-based feature extraction, GAN-based augmentation, and MGA-CSG optimization contributes to the model's high performance. This approach holds significant potential in improving healthcare management and providing timely interventions for patients with retinal diseases.

Classification of Progressive Wear on a Multi-Directional Pin-on-Disc Tribometer Simulating Conditions in Human Joints-UHMWPE against CoCrMo Using Acoustic Emission and Machine Learning

Human joint prostheses experience wear failure due to the complex interactions between Ultra-High-Molecular-Weight Polyethylene (UHMWPE) and Cobalt-Chromium-Molybdenum (CoCrMo). This study uses the wear classification to investigate the gradual and progressive abrasive wear mechanisms in UHMWPE. Pin-on-disc tests were conducted under simulated in vivo conditions, monitoring wear using Acoustic Emission (AE). Two Machine Learning (ML) frameworks were employed for wear classification: manual feature extraction with ML classifiers and a contrastive learning-based Convolutional Neural Network (CNN) with ML classifiers. The CNN-based feature extraction approach achieved superior classification performance (94% to 96%) compared to manual feature extraction (81% to 89%). The ML techniques enable accurate wear classification, aiding in understanding surface states and early failure detection. Real-time monitoring using AE sensors shows promise for interventions and improving prosthetic joint design.

A brain-inspired approach for SAR-to-optical image translation based on diffusion models.

Synthetic Aperture Radar (SAR) plays a crucial role in all-weather and all-day Earth observation owing to its distinctive imaging mechanism. However, interpreting SAR images is not as intuitive as optical images. Therefore, to make SAR images consistent with human cognitive habits and assist inexperienced people in interpreting SAR images, a generative model is needed to realize the translation from SAR images to optical ones. In this work, inspired by the processing of the human brain in painting, a novel conditional image-to-image translation framework is proposed for SAR to optical image translation based on the diffusion model. Firstly, considering the limited performance of existing CNN-based feature extraction modules, the model draws insights from the self-attention and long-skip connection mechanisms to enhance feature extraction capabilities, which are aligned more closely with the memory paradigm observed in the functioning of human brain neurons. Secondly, addressing the scarcity of SAR-optical image pairs, data augmentation that does not leak the augmented mode into the generated mode is designed to optimize data efficiency. The proposed SAR-to-optical image translation method is thoroughly evaluated using the SAR2Opt dataset. Experimental results demonstrate its capacity to synthesize high-fidelity optical images without introducing blurriness.

Detection of Cocoa Leaf Diseases Using the CNN-Based Feature Extractor and XGBOOST Classifier

Detection of Cocoa Leaf Diseases Using the CNN-Based Feature Extractor and XGBOOST Classifier

Early Detection of Brain Tumor from MRI Images Using Different Machine Learning Techniques

Effective treatment of brain tumours depends on early identification of tumour tissues. Categorizing tumors is essential for its early identification. In order to split and categorise brain tumours using MRI images, various machine learning algorithms are examined. Preprocessing methods including image augmentation, rotation flipping, and adaptive deformation are used on the BRATs dataset to improve the data quality. Convolutional neural network (CNN) feature extraction techniques like VGG19 and InceptionV3 are applied. K-Nearest Neighbours (KNN) and logistic regression techniques are used to categorise the features. The proposed models’ accuracy for Logistic Regression, KNN, VGG19, Inception V3 and VGG19 with augmentation is 84.66%, 90.68%, 92.17%, 91.56% and 95.43% respectively. These findings imply that CNN-based feature extraction and picture augmentation techniques greatly improves the accuracy of brain tumour identification, Overall thecInception V3 with augmentation produces the best results.

Skin Lesion Classification Using Convolutional Neural Network and ABCD Rule

Skin cancer, which can occur in any part of the human skin, is one of the common and serious types of cancer. Accurate diagnosis and segmentation of lesions are crutial to the early diagnosis. Computer-aided diagnosis make important contributions to help doctors in the diagnosis of cancer from skin images. The most important factor for such systems to reveal the accurate results is the correct feature extraction. In this study, a model for the classification of seven types of skin lesions was developed by combining the features of CNN-based feature extraction and the ABCD rule, which is widely used in the clinic. The model was evaluated on HAM10000 well-known dataset. The classification results obtained with different combinations of features and machine learning algorithms were compared. According to the results, the best classification accuracy was obtained with the Cosine Similarity Classifier with 96.4% when the features determined by CNN and the features in the ABCD rule were used together.

Learning Heavily-Degraded Prior for Underwater Object Detection

Underwater object detection suffers from low detection performance because the distance and wavelength dependent imaging process yield evident image quality degradations such as haze-like effects, low visibility, and color distortions. Therefore, we commit to resolving the issue of underwater object detection with compounded environmental degradations. Typical approaches attempt to develop sophisticated deep architecture to generate high-quality images or features. However, these methods are only work for limited ranges because imaging factors are either unstable, too sensitive, or compounded. Unlike these approaches catering for high-quality images or features, this paper seeks transferable prior knowledge from detector-friendly images. The prior guides detectors removing degradations that interfere with detection. It is based on statistical observations that, the heavily degraded regions of detector-friendly (DFUI) and underwater images have evident feature distribution gaps while the lightly degraded regions of them overlap each other. Therefore, we propose a residual feature transference module (RFTM) to learn a mapping between deep representations of the heavily degraded patches of DFUI- and underwater- images, and make the mapping as a heavily degraded prior (HDP) for underwater detection. Since the statistical properties are independent to image content, HDP can be learned without the supervision of semantic labels and plugged into popular CNN-based feature extraction networks to improve their performance on underwater object detection. Without bells and whistles, evaluations on URPC2020 and UODD show that our methods outperform CNN-based detectors by a large margin. Our method with higher speeds and less parameters still performs better than transformer-based detectors. Our code and DFUI dataset can be found in https://github.com/xiaoDetection/Learning-Heavily-Degraed-Prior.

A CNN based multifaceted signal processing framework for heart rate proctoring using Millimeter wave radar ballistocardiography

The recent pandemic has refocused the medical world’s attention on the diagnostic techniques associated with cardiovascular disease. Heart rate provides a real-time snapshot of cardiovascular health. A more precise heart rate reading enables a better understanding of cardiac muscle activity. Although many existing diagnostic techniques are approaching the limits of perfection, there remains potential for further development. In this paper, we propose MIBINET, a novel multifaceted approach for real-time proctoring of heart rate from Millimeter wave (mm-wave) radar ballistocardiography signals via inter-beat-interval (IBI) using a convolutional neural NETwork (CNN). The central theme of our approach is to synergize the feature extraction capabilities of CNN with novel signal processing techniques, resulting in enhanced estimation accuracy while simultaneously reducing computational complexity. This proposed network can be used in hospitals, homes, and passenger vehicles due to its lightweight and contactless properties. It employs classical signal processing prior to fitting the data into the network. Although MIBINET is primarily designed to work on mm-wave signals, it is found equally effective on signals of various modalities such as PCG, ECG, and PPG. Our approach outperforms state-of-the-art techniques by more than 5% in inter-beat-interval (IBI) estimation accuracy. The architecture achieves a 98.73% correlation coefficient and a 20.69 ms Root-Mean-Square Error (RMSE) over 11 different test subjects. The paper contributes by being the first to apply CNN-based feature extraction in concert with unique signal processing strategies to mm-wave radar data for heart rate monitoring. Our methodology also introduces a synthetic IBI augmentation technique, custom loss function, and novel post-processing methods, all contributing to the robust performance of the model in various settings and modalities.

Hybrid generative model for grading the severity of diabetic retinopathy images

ABSTRACT One of the common eye conditions affecting patients with diabetes is diabetic retinopathy (DR). It is characterised by the progressive impairment to the blood vessels with the increase of glucose level in the blood. The grading efficiency still finds challenging because of the existence of intra-class variations and imbalanced data distributions on the retinal images. Traditional machine learning techniques utilise hand-engineered features for classification of the affected retinal images. As convolutional neural network produces better image classification accuracy in many medical images, this work utilises the CNN-based feature extraction method. This feature has been used to build Gaussian mixture model (GMM) for each class that maps the CNN features to log-likelihood dimensional vector spaces. Since the Gaussian mixture model can be realised as a mixture of both parametric and nonparametric density models and has their flexibility in capturing different data distributions, probabilistic outputs, interpretability, efficient parameter estimation, and robustness to outliers, the proposed model aimed to obtain and provide a smooth approximation of the underlying distribution of features for training the model. Then these vector spaces are trained by the SVM classifier. Experimental results illustrate the efficacy of the proposed model with accuracy 86.3% and 89.1%, respectively.

Transformer vehicle re-identification of intelligent transportation system under carbon neutral target

Vehicle re-identification technology plays a pivotal role in intelligent transportation systems, contributing significantly to improved traffic scheduling and reduced net-zero emissions. It holds immense potential for advancing carbon neutrality efforts. By accurately recognizing and tracking vehicles, this technology optimizes traffic flow, enhances energy efficiency, reduces carbon emissions, and fosters sustainable transportation development. However, prevailing CNN-based feature extraction methods often lack the capability to associate global features, while Transformer-based approaches tend to overlook crucial local feature differences. Effectively mining local features and integrating them with global features are critical aspects of successful vehicle re-identification techniques. To tackle this challenge, we present a novel approach, the Locally Significant Feature Rearrangement (LSFR) module, based on the Swin Transformer model. This module enhances the learning of local features by initially focusing on specific regions using a single-shot multi-box detector (SSD) and employing an adaptive attention learning mechanism to highlight fine-grained local features. Subsequently, the salient local features are rearranged to prioritize their importance, and their embedding layers are fused with global features in a recombined manner. Our experimental results on the VeRi-776 and VehicleID benchmark datasets validate the effectiveness of our proposed method as a reliable technique to support intelligent transportation systems and drive progress towards traffic carbon neutralization.