Search

IntroductionWhen it comes to interpreting visual input, intelligent systems make use of contextual scene learning, which significantly improves both resilience and context awareness. The management of enormous amounts of data is a driving force behind the growing interest in computational frameworks, particularly in the context of autonomous cars.MethodThe purpose of this study is to introduce a novel approach known as Deep Fused Networks (DFN), which improves contextual scene comprehension by merging multi-object detection and semantic analysis.ResultsTo enhance accuracy and comprehension in complex situations, DFN makes use of a combination of deep learning and fusion techniques. With a minimum gain of 6.4% in accuracy for the SUN-RGB-D dataset and 3.6% for the NYU-Dv2 dataset.DiscussionFindings demonstrate considerable enhancements in object detection and semantic analysis when compared to the methodologies that are currently being utilized.

IntroductionAdvanced traffic monitoring systems face significant challenges in vehicle detection and classification. Conventional methods often require substantial computational resources and struggle to adapt to diverse data collection methods.MethodsThis research introduces an innovative technique for classifying and recognizing vehicles in aerial image sequences. The proposed model encompasses several phases, starting with image enhancement through noise reduction and Contrast Limited Adaptive Histogram Equalization (CLAHE). Following this, contour-based segmentation and Fuzzy C-means segmentation (FCM) are applied to identify foreground objects. Vehicle detection and identification are performed using EfficientDet. For feature extraction, Accelerated KAZE (AKAZE), Oriented FAST and Rotated BRIEF (ORB), and Scale Invariant Feature Transform (SIFT) are utilized. Object classification is achieved through a Convolutional Neural Network (CNN) and ResNet Residual Network.ResultsThe proposed method demonstrates improved performance over previous approaches. Experiments on datasets including Vehicle Aerial Imagery from a Drone (VAID) and Unmanned Aerial Vehicle Intruder Dataset (UAVID) reveal that the model achieves an accuracy of 96.6% on UAVID and 97% on VAID.DiscussionThe results indicate that the proposed model significantly enhances vehicle detection and classification in aerial images, surpassing existing methods and offering notable improvements for traffic monitoring systems.

IntroductionUnmanned aerial vehicles (UAVs) are widely used in various computer vision applications, especially in intelligent traffic monitoring, as they are agile and simplify operations while boosting efficiency. However, automating these procedures is still a significant challenge due to the difficulty of extracting foreground (vehicle) information from complex traffic scenes.MethodsThis paper presents a unique method for autonomous vehicle surveillance that uses FCM to segment aerial images. YOLOv8, which is known for its ability to detect tiny objects, is then used to detect vehicles. Additionally, a system that utilizes ORB features is employed to support vehicle recognition, assignment, and recovery across picture frames. Vehicle tracking is accomplished using DeepSORT, which elegantly combines Kalman filtering with deep learning to achieve precise results.ResultsOur proposed model demonstrates remarkable performance in vehicle identification and tracking with precision of 0.86 and 0.84 on the VEDAI and SRTID datasets, respectively, for vehicle detection.DiscussionFor vehicle tracking, the model achieves accuracies of 0.89 and 0.85 on the VEDAI and SRTID datasets, respectively.

IntroductionFalls are a major cause of accidents that can lead to serious injuries, especially among geriatric populations worldwide. Ensuring constant supervision in hospitals or smart environments while maintaining comfort and privacy is practically impossible. Therefore, fall detection has become a significant area of research, particularly with the use of multimodal sensors. The lack of efficient techniques for automatic fall detection hampers the creation of effective preventative tools capable of identifying falls during physical exercise in long-term care environments. The primary goal of this article is to examine the benefits of using multimodal sensors to enhance the precision of fall detection systems.MethodsThe proposed paper combines time–frequency features of inertial sensors with skeleton-based modeling of depth sensors to extract features. These multimodal sensors are then integrated using a fusion technique. Optimization and a modified K-Ary classifier are subsequently applied to the resultant fused data.ResultsThe suggested model achieved an accuracy of 97.97% on the UP-Fall Detection dataset and 97.89% on the UR-Fall Detection dataset.DiscussionThis indicates that the proposed model outperforms state-of-the-art classification results. Additionally, the proposed model can be utilized as an IoT-based solution, effectively promoting the development of tools to prevent fall-related injuries.

IntroductionHand gestures are an effective communication tool that may convey a wealth of information in a variety of sectors, including medical and education. E-learning has grown significantly in the last several years and is now an essential resource for many businesses. Still, there has not been much research conducted on the use of hand gestures in e-learning. Similar to this, gestures are frequently used by medical professionals to help with diagnosis and treatment.MethodWe aim to improve the way instructors, students, and medical professionals receive information by introducing a dynamic method for hand gesture monitoring and recognition. Six modules make up our approach: video-to-frame conversion, preprocessing for quality enhancement, hand skeleton mapping with single shot multibox detector (SSMD) tracking, hand detection using background modeling and convolutional neural network (CNN) bounding box technique, feature extraction using point-based and full-hand coverage techniques, and optimization using a population-based incremental learning algorithm. Next, a 1D CNN classifier is used to identify hand motions.ResultsAfter a lot of trial and error, we were able to obtain a hand tracking accuracy of 83.71% and 85.71% over the Indian Sign Language and WLASL datasets, respectively. Our findings show how well our method works to recognize hand motions.DiscussionTeachers, students, and medical professionals can all efficiently transmit and comprehend information by utilizing our suggested system. The obtained accuracy rates highlight how our method might improve communication and make information exchange easier in various domains.

Honeybees play a vital role in preservation of an healthy environment. Bees not only provide pollination services but also produce honey, beeswax, and royal jelly. Beekeeping has a rich history and substantial economic potential worldwide, but swarming remains a crucial challenge for maintaining profitability. Swarming, a typical colony reproductive process in honeybees, significantly impacts beekeepers profitability by lowering the number of bees in hives and thus effecting honey production. Monitoring of these beehives is therefore of paramount importance to keep an eye on their irregular behavior. Swarm prediction can be done by visually inspecting hives, monitoring temperature, or analyzing acoustic features with machine learning. Acoustic monitoring is instrumental in detecting changes in colony behavior since it overcomes the constraints of visual inspections and is not affected by external factors like temperature. In this paper, we aim to evaluate various state-of-the-art machine learning and deep learning models for swarm prediction by studying wave plot features, Mel Spectrogram, and Melfrequency Cepstral coefficients (MFCC). We use Naive Bayes, K-nearest Neighbors (KNN), and Support Vector Machines (SVM) as machine learning models and Convolution Neural Networks (CNN), Long Short Term Memory (LSTM), and Transformers as deep learning models for comparison purposes. We apply these models on a well-known honey bees audio dataset provided by the NU-hive project and consider classification metrics such as accuracy, precision, recall, and F1 score for the comparative evaluation of our models. Our evaluation demonstrates SVM as the best-performing machine learning algorithm. In particular, SVM with Mel Spectrogram as input data, achieved an accuracy of around 97%. On the other hand, CNN outperformed all the models and achieved an accuracy of 99%, using MFCC features as input data. As a result of these encouraging outcomes, we understand that our results can help the researchers to choose which AI model is more suitable for them to design beehive monitoring systems for accurate identification of abnormal situations in beehives.

Object Classification in Remote Sensing Imagery holds paramount importance for extracting meaningful insights from complex aerial scenes. Conventional methods encounter challenges in achieving precision amidst diverse landscape features. This paper introduces an innovative hybrid model to enhance the accuracy of remote sensing multi-object classification. Incorporating a feature-level fusion approach inspired by successful methods, we leverage Adoptive Fuzzy C-means segmentation for precise object classification and Conditional Random Field labeling. Our model excels in capturing diverse features within remote sensing images using multiple feature extraction methods. The distinctive feature of our methodology lies in the thoughtful incorporation of a Deep Belief Network. Through rigorous experimental evaluations on two standard datasets, our proposed system demonstrates exceptional performance, emphasizing its significant potential for advancing methodologies in remote sensing multi-object classification. This tactful integration results in substantial improvements, yielding high accuracies of 97.24% (UCM) and 96.84% (RESISC45). The proposed model is methodologically novel and effective solution for advancing remote sensing image classification.

In computer vision and image processing, vehicle detection and tracking in complicated aerial images have become important subjects. The need for automated systems that can precisely detect and track vehicles in aerial image data is growing due to the abundance of data coming from numerous sources, including drones and satellites. This study introduces a new method for lane extraction that relies on centroid tracking and template matching, followed by co-registration and geo-referencing. Our approach offers robust vehicle detection and tracking over a range of sizes and positions in complicated backgrounds, while also efficiently segmenting the region of interest. Our suggested method, which makes use of machine learning and feature extraction techniques, shows excellent precision and effectiveness when it comes to detecting and tracking vehicles in complex aerial images. This finding has important implications for traffic management and urban planning, going beyond computer vision and image processing. Our technology has the potential to transform traffic management procedures by making it simpler to detect traffic bottlenecks and monitor traffic flow. Furthermore, our method can help identify damaged vehicles in disaster response scenarios, which will help prioritize rescue and recovery activities. All things considered, our suggested approach is a significant addition to the domains of computer vision and image processing, with a broad range of uses in traffic control, urban planning, and disaster management.

Advancements in smartphone sensor technologies have significantly enriched the field of human activity recognition, facilitating a wide array of applications from health monitoring to personal navigation. This study utilized such advancements to explore human locomotion and localization recognition using data from accelerometers, microphones, gyroscopes, magnetometers, and GPS, applying Deep Polynomial Neural Networks (DPNN) and Multilayer Perceptron (MLP) across three datasets: the Continuous In-The-Wild Smart Watch Activity Dataset, the Huawei Locomotion Dataset, and the Extra Sensory Dataset. We employ two distinct approaches for activity recognition: Deep Polynomial Neural Networks (DPNN) for deep learning-based feature extraction and Multilayer Perceptron (MLP) with manual feature extraction techniques, including Linear Predictive Coding Cepstral Coefficients (LPCC), step length, signal magnitude area, spectral, and sound features. Through rigorous experimentation, we achieved remarkable accuracy in recognizing both locomotion and localization activities, with DPNN consistently outperforming MLP in terms of accuracy. Specifically, for the Continuous In-The-Wild Dataset, DPNN achieved a 93% accuracy rate for localization activities and 95% for locomotion activities, while MLP recorded 86% and 91% in the respective categories. Similarly, on the Huawei Locomotion Dataset, DPNN attained 95% accuracy for localization and 97% for locomotion, with MLP achieving 88% and 91%, respectively. Furthermore, the application of these models to the Extra Sensory Dataset yielded 92% accuracy for both localization and locomotion activities with DPNN, and 90% and 89% with MLP. In our study, we observed that in terms of accuracy, DPNN emerges as the clear winner; however, it is computationally expensive. Conversely, MLP, while being less accurate, stands out for its computational efficiency. This study not only highlights the effectiveness of incorporating advanced machine learning techniques in interpreting sensor data but also emphasizes the trade-offs between computational demands and accuracy in the domain of human activity recognition. Through our comprehensive analysis, we contribute valuable insights into the potential of smartphone sensors in enhancing activity recognition systems, paving the way for future innovations in mobile sensing technology.

In the field of traffic surveillance systems, where effective traffic management and safety are the primary concerns, vehicle detection and tracking play an important role. Low brightness, low contrast, and noise are issues with low-light environments that result from poor lighting or insufficient exposure. In this paper, we proposed a vehicle detection and tracking model based on the aerial image captured during nighttime. Before object detection, we performed fogging and image enhancement using MIRNet architecture. After pre-processing, YOLOv5 was used to locate each vehicle position in the image. Each detected vehicle was subjected to a Scale-Invariant Feature Transform (SIFT) feature extraction algorithm to assign a unique identifier to track multiple vehicles in the image frames. To get the best possible location of vehicles in the succeeding frames templates were extracted and template matching was performed. The proposed model achieves a precision score of 0.924 for detection and 0.861 for tracking with the Unmanned Aerial Vehicle Benchmark Object Detection and Tracking (UAVDT) dataset, 0.904 for detection, and 0.833 for tracking with the Vision Meets Drone Single Object-Tracking (VisDrone) dataset.

Object recognition along with classification are necessary for many applications, such as surveillance systems, car plate recognition, traffic monitoring, and face detection. Unlike existing approaches, ours incorporates a wide range of important factors to improve recognition precision. The primary phase in the image accumulating process is preprocessing, when semantic segmentation proves its usefulness by accurately defining the physical borders of specific objects inside an image in addition to recognizing them. This paper presents a novel approach to accurate object recognition. Segmentation incorporates previously identified homologous and related groups after employing the K-means clustering technique to group analogous colors and spatial patterns. Convolutional Neural Network (CNN) technology is ultimately used to identify objects in different environmental circumstances. Performance metrics like as F1 Score=0.948, Precision = 0.968, and Recall=0.932 for MSRC and F1 Score=0.921, Precision = 0.951, and Recall=0.891 for Caltech 101 and F1 Score=0.847, Precision = 0.879, and Recall=0.827 over Pascal Voc 2012 demonstrate the efficiency of our strategy. The efficacy of the suggested method is evaluated using multiple benchmark datasets, MSRC-v2, Caltech 101 and Pascal Voc 2012, yielding recognition accuracies of 92.25%, 91.91% and 93.50% respectively, when tested against the Microsoft Research Cambridge (MSRC), California Institute of Technology 101 Object Categories (Caltech 101) and Pascal Voc 2012 datasets.

Systems must be capable of detecting and tracking autonomous vehicles for intelligent management and control of transportation. Even though several methods are used to create intelligent systems for traffic monitoring, this article explains how to detect and track vehicles using pixel labeling and particle filter algorithms. We suggested a novel technique that segments the image using image segmentation to retrieve the foreground objects. We have divided our proposed model into the following steps: at first, geo-referencing is used to find the exact location; secondly, the images are denoised by using pre-processing; image segmentation is used to separate the background from the foreground; multiple objection detection is performed using the random forest to classify different objects; vehicles are detected through a method called template matching; after this, the vehicles are counted using histogram of oriented gradients (HOG); and after counting, the tracking of vehicles is obtained using particle filter; lastly, the trajectories are predicted by comparing the rectangular centroid of each car against the frame number and using it as a time stamp reference, the last match that the tracking algorithm obtained for each vehicle was recorded and used to estimate the trajectories. Our model outperforms current traffic monitoring approaches in terms of detection and tracking accuracy by 0.87 and 0.92, respectively using the Aerial Car dataset and 0.84 and 0.88, respectively, using the AU-AIR datasets. Vehicle recognition, traffic density detection, traffic flow analysis, and pedestrian route generation are all possible uses for the proposed system. It can transform traffic management and improve overall road safety due to its powerful algorithms and cutting-edge technologies. The system’s adaptability makes it a significant asset in modern transportation, from optimizing signal timing to improving pedestrian navigation.

Hemoglobin is the iron containing protein in red blood cells which carries oxygen from lungs to rest of the body tissues. Accurate measurement of hemoglobin is essential for diagnosing anemia, a condition characterized by a deficiency of red blood cells. This measurement is particularly vital before initiating blood transfusions for thalassemia patients. Non-invasive estimation of hemoglobin levels can be achieved through photoplethysmography (PPG)-based methods. PPG is an optical method to measure blood volume changes in successive heart beats. PPG signals can be obtained from fingertip videos using a light source and a photodetector. SmartphonePPG utilizes a smartphone’s flashlight as a light source and its camera as a photodetector to acquire PPG signals, offering an affordable and portable point-of-care tool. Despite the ubiquity of smartphones, signals from their cameras often contain noise, making feature selection from PPG characteristics challenging. While PPG-based methods are invaluable, the lack of real-world datasets poses a significant challenge in maximizing the benefits of PPG technology. In this paper, we introduce a dataset comprising 1-minute fingertip video recordings from 150 anemic patients, obtained using a smartphone’s camera. The dataset, publicly accessible for research purposes (https://forms.gle/LB4qn81ZMqEuy3V27), covers an age range of 6 months to 32 years, with diverse hemoglobin values (4.3 gm/dL - 12.4 gm/dL). Utilizing this dataset, we propose a deep learning-based technique employing the ResNet-18 architecture to estimate hemoglobin levels. This approach eliminates the need for manual feature extraction and selection from PPG signals, overcoming a limitation in existing smartphone PPG-based hemoglobin estimation systems. Our model achieves a hemoglobin level estimation with an RMSE of 0.81-1.39 when compared with the gold standard laboratory method, Complete Blood Count (CBC) test reports. In contrast, HemaApp, a state-of-the-art research utilizing a machine learning-based classifier (SVM), yields an RMSE of 1.7 on our dataset. The accuracy and simplicity of our model position it as a promising alternative to existing non-invasive hemoglobin level estimation methods.

Parkinson’s disease affects bodily functions and there is a growing need for advanced solutions to offer therapeutic advice to patients. A framework using artificial intelligence and machine learning techniques has been proposed to address this. The proposed system employs a combination of RGB, inertial, and depth sensors data. The inertial signals have been filtered using a notch filter to obtain the optimal wearable sensor data by examining the upper and lower cutoff frequencies. Multiple features have been calculated, including mel frequency cepstral coefficients (MFCC), statistical features and Gaussian mixture model (GMM) features. On the other hand, silhouettes have been extracted from RGB and depth images, and four crucial parameters have been employed to gauge the level of accuracy with which patients with neurological disorders performed their activities, including the angle formation between the hands and lower/upper half of the body and center of the body, and the angle between the hands of the silhouette. The resulting features have been fused and classified using principle component analysis and a reweighted genetic algorithm. Evaluation using cross-validation on mRI (multi-modal 3d human pose estimation dataset using mmwave, RGB-D, and inertial sensors) and MHEALTH (mobile health) datasets showed a recognition accuracy rate of 97.29% and a 97.94%. The study highlights the need for more datasets to address challenges in rehabilitation using human activity recognition with multi-modal sensors.

Machines need to be able to recognize and understand complex visual surroundings to function at their best in a variety of contexts. Here, we address the difficult problem of multi-object recognition to obtain a sophisticated knowledge of complex visual environments, tackling issues such as size, occlusion, fluctuations in object traits, and complicated backdrops. Our contribution is to provide novel methods (Gaussian mixture model and mean-shift algorithms) for inferring multiple object segmentation in complicated visuals, introducing a unique multiclass object classification strategy utilizing benchmark datasets. Notably, by utilizing local appearance, texture, and geometry characteristics, our technique considerably improves classification accuracy by integrating a Multi-Layer Perceptron (MLP) with area signatures and local descriptors. By facilitating accurate object matching and identification based on local appearance, texture, and geometric features, local descriptors are essential for collecting particular information and regions of interest in images. When compared to state-of-the-art methods, empirical validation on MSRC and Corel 10k datasets shows better performance, especially when managing object occlusion problems. With an accuracy of 90.6% and 89.69%, respectively, our suggested system performs better than industry standards for multi-object classification on both datasets, highlighting the significant progress our method makes to the area of multiclass object classification in challenging visual contexts.

Knowledge graph (KG) summarization facilitates efficient information retrieval for exploring complex structural data. For fast information retrieval, it requires processing on redundant data. However, it necessitates the completion of information in a summary graph. It also saves computational time during data retrieval, storage space, in-memory visualization, and preserving structure after summarization. State-of-the-art approaches summarize a given $KG$ by preserving its structure at the cost of information loss. Additionally, the approaches not preserving the underlying structure, compromise the summarization ratio by focusing only on the compression of specific regions. In this way, these approaches either miss preserving the original facts or the wrong prediction of inferred information. To solve these problems, we present a novel framework for generating a lossless summary by preserving the structure through super signatures and their corresponding corrections. The proposed approach summarizes only the naturally overlapped instances while maintaining its information and preserving the underlying Resource Description Framework $RDF$ graph. The resultant summary is composed of triples with positive, negative, and star corrections that are optimized by the smart calling of two novel functions namely $merge$ and $disperse$ . To evaluate the effectiveness of our proposed approach, we perform experiments on nine publicly available real-world knowledge graphs and obtain a better summarization ratio than state-of-the-art approaches by a margin of 10% to 30% with achieving its completeness, correctness, and compactness. In this way, the retrieval of common events and groups by queries is accelerated in the resultant graph.

Human activity recognition (HAR) plays a pivotal role in various domains, including healthcare, sports, robotics, and security. With the growing popularity of wearable devices, particularly Inertial Measurement Units (IMUs) and Ambient sensors, researchers and engineers have sought to take advantage of these advances to accurately and efficiently detect and classify human activities. This research paper presents an advanced methodology for human activity and localization recognition, utilizing smartphone IMU, Ambient, GPS, and Audio sensor data from two public benchmark datasets: the Opportunity dataset and the Extrasensory dataset. The Opportunity dataset was collected from 12 subjects participating in a range of daily activities, and it captures data from various body-worn and object-associated sensors. The Extrasensory dataset features data from 60 participants, including thousands of data samples from smartphone and smartwatch sensors, labeled with a wide array of human activities. Our study incorporates novel feature extraction techniques for signal, GPS, and audio sensor data. Specifically, for localization, GPS, audio, and IMU sensors are utilized, while IMU and Ambient sensors are employed for locomotion activity recognition. To achieve accurate activity classification, state-of-the-art deep learning techniques, such as convolutional neural networks (CNN) and long short-term memory (LSTM), have been explored. For indoor/outdoor activities, CNNs are applied, while LSTMs are utilized for locomotion activity recognition. The proposed system has been evaluated using the k-fold cross-validation method, achieving accuracy rates of 97% and 89% for locomotion activity over the Opportunity and Extrasensory datasets, respectively, and 96% for indoor/outdoor activity over the Extrasensory dataset. These results highlight the efficiency of our methodology in accurately detecting various human activities, showing its potential for real-world applications. Moreover, the research paper introduces a hybrid system that combines machine learning and deep learning features, enhancing activity recognition performance by leveraging the strengths of both approaches.

The acquisition, processing, mining, and visualization of sensory data for knowledge discovery and decision support has recently been a popular area of research and exploration. Its usefulness is paramount because of its relationship to the continuous involvement in the improvement of healthcare and other related disciplines. As a result of this, a huge amount of data have been collected and analyzed. These data are made available for the research community in various shapes and formats; their representation and study in the form of graphs or networks is also an area of research which many scholars are focused on. However, the large size of such graph datasets poses challenges in data mining and visualization. For example, knowledge discovery from the Bio–Mouse–Gene dataset, which has over 43 thousand nodes and 14.5 million edges, is a non-trivial job. In this regard, summarizing the large graphs provided is a useful alternative. Graph summarization aims to provide the efficient analysis of such complex and large-sized data; hence, it is a beneficial approach. During summarization, all the nodes that have similar structural properties are merged together. In doing so, traditional methods often overlook the importance of personalizing the summary, which would be helpful in highlighting certain targeted nodes. Personalized or context-specific scenarios require a more tailored approach for accurately capturing distinct patterns and trends. Hence, the concept of personalized graph summarization aims to acquire a concise depiction of the graph, emphasizing connections that are closer in proximity to a specific set of given target nodes. In this paper, we present a faster algorithm for the personalized graph summarization (PGS) problem, named IPGS; this has been designed to facilitate enhanced and effective data mining and visualization of datasets from various domains, including biosensors. Our objective is to obtain a similar compression ratio as the one provided by the state-of-the-art PGS algorithm, but in a faster manner. To achieve this, we improve the execution time of the current state-of-the-art approach by using weighted, locality-sensitive hashing, through experiments on eight large publicly available datasets. The experiments demonstrate the effectiveness and scalability of IPGS while providing a similar compression ratio to the state-of-the-art approach. In this way, our research contributes to the study and analysis of sensory datasets through the perspective of graph summarization. We have also presented a detailed study on the Bio–Mouse–Gene dataset, which was conducted to investigate the effectiveness of graph summarization in the domain of biosensors.

The domain of human locomotion identification through smartphone sensors is witnessing rapid expansion within the realm of research. This domain boasts significant potential across various sectors, including healthcare, sports, security systems, home automation, and real-time location tracking. Despite the considerable volume of existing research, the greater portion of it has primarily concentrated on locomotion activities. Comparatively less emphasis has been placed on the recognition of human localization patterns. In the current study, we introduce a system by facilitating the recognition of both human physical and location-based patterns. This system utilizes the capabilities of smartphone sensors to achieve its objectives. Our goal is to develop a system that can accurately identify different human physical and localization activities, such as walking, running, jumping, indoor, and outdoor activities. To achieve this, we perform preprocessing on the raw sensor data using a Butterworth filter for inertial sensors and a Median Filter for Global Positioning System (GPS) and then applying Hamming windowing techniques to segment the filtered data. We then extract features from the raw inertial and GPS sensors and select relevant features using the variance threshold feature selection method. The extrasensory dataset exhibits an imbalanced number of samples for certain activities. To address this issue, the permutation-based data augmentation technique is employed. The augmented features are optimized using the Yeo–Johnson power transformation algorithm before being sent to a multi-layer perceptron for classification. We evaluate our system using the K-fold cross-validation technique. The datasets used in this study are the Extrasensory and Sussex Huawei Locomotion (SHL), which contain both physical and localization activities. Our experiments demonstrate that our system achieves high accuracy with 96% and 94% over Extrasensory and SHL in physical activities and 94% and 91% over Extrasensory and SHL in the location-based activities, outperforming previous state-of-the-art methods in recognizing both types of activities.

Advancements in sensing technology have expanded the capabilities of both wearable devices and smartphones, which are now commonly equipped with inertial sensors such as accelerometers and gyroscopes. Initially, these sensors were used for device feature advancement, but now, they can be used for a variety of applications. Human activity recognition (HAR) is an interesting research area that can be used for many applications like health monitoring, sports, fitness, medical purposes, etc. In this research, we designed an advanced system that recognizes different human locomotion and localization activities. The data were collected from raw sensors that contain noise. In the first step, we detail our noise removal process, which employs a Chebyshev type 1 filter to clean the raw sensor data, and then the signal is segmented by utilizing Hamming windows. After that, features were extracted for different sensors. To select the best feature for the system, the recursive feature elimination method was used. We then used SMOTE data augmentation techniques to solve the imbalanced nature of the Extrasensory dataset. Finally, the augmented and balanced data were sent to a long short-term memory (LSTM) deep learning classifier for classification. The datasets used in this research were Real-World Har, Real-Life Har, and Extrasensory. The presented system achieved 89% for Real-Life Har, 85% for Real-World Har, and 95% for the Extrasensory dataset. The proposed system outperforms the available state-of-the-art methods.

Multiple Internet of Healthcare Things (IoHT)-based devices have been utilized as sensing methodologies for human locomotion decoding to aid in applications related to e-healthcare. Different measurement conditions affect the daily routine monitoring, including the sensor type, wearing style, data retrieval method, and processing model. Currently, several models are present in this domain that include a variety of techniques for pre-processing, descriptor extraction, and reduction, along with the classification of data captured from multiple sensors. However, such models consisting of multiple subject-based data using different techniques may degrade the accuracy rate of locomotion decoding. Therefore, this study proposes a deep neural network model that not only applies the state-of-the-art Quaternion-based filtration technique for motion and ambient data along with background subtraction and skeleton modeling for video-based data, but also learns important descriptors from novel graph-based representations and Gaussian Markov random-field mechanisms. Due to the non-linear nature of data, these descriptors are further utilized to extract the codebook via the Gaussian mixture regression model. Furthermore, the codebook is provided to the recurrent neural network to classify the activities for the locomotion-decoding system. We show the validity of the proposed model across two publicly available data sampling strategies, namely, the HWU-USP and LARa datasets. The proposed model is significantly improved over previous systems, as it achieved 82.22% and 82.50% for the HWU-USP and LARa datasets, respectively. The proposed IoHT-based locomotion-decoding model is useful for unobtrusive human activity recognition over extended periods in e-healthcare facilities.