Smart cushion-based non-invasive mental fatigue assessment of construction equipment operators: A feasible study

Advanced Engineering Informatics, Год журнала: 2023, Номер 58, С. 102134 - 102134

Опубликована: Авг. 23, 2023

Язык: Английский

---

Early diagnosis of respiratory system diseases (RSD) using deep convolutional neural networks

Journal of Ambient Intelligence and Humanized Computing, Год журнала: 2023, Номер 14(9), С. 12273 - 12283

Опубликована: Июль 7, 2023

Язык: Английский

---

A machine-learning Approach for Stress Detection Using Wearable Sensors in Free-living Environments

medRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Апрель 30, 2024

Abstract Stress is a psychological condition due to the body’s response challenging situation. If person exposed prolonged periods and various forms of stress, their physical mental health can be negatively affected, leading chronic problems. It important detect stress in its initial stages prevent stress-related issues. Thus, there must alternative effective solutions for spontaneous monitoring. Wearable sensors are one most prominent solutions, given capacity collect data continuously real-time. sensors, among others, have been widely used bridge existing gaps monitoring thanks non-intrusive nature. Besides, they monitor vital signs, e.g., heart rate activity. Yet, works focused on acquired controlled settings. To this end, our study aims propose machine learning-based approach detecting onsets free-living environment using wearable sensors. The authors utilized SWEET dataset collected from 240 subjects via electrocardiography (ECG), skin temperature (ST), conductance (SC). In work, four learning models were tested set consisting subjects, namely K-Nearest Neighbors (KNN), Support vector classification (SVC), Decision Tree (DT), Random Forest (RF). These trained scenarios. Neighbor (KNN) model had highest accuracy 98%, while other also performed satisfactorily.

Язык: Английский

---

Security enhancement of the access control scheme in IoMT applications based on fuzzy logic processing and lightweight encryption

Complex & Intelligent Systems, Год журнала: 2023, Номер 10(1), С. 435 - 454

Опубликована: Июль 25, 2023

Abstract Security of Internet-of-Medical-Things (IoMT) networks has evolved as a vital issue in recent years. The IoMT are designed to link patients with caregivers. All reports, data, and medical signals transferred over these networks. Hence, require robust secure access strategies for send their data or reports. hacking may lead harmful effects on patients. One the vulnerable points is point. Access could be performed biometrics. popular biometric traits this purpose biomedical such Electrocardiogram (ECG) signals, they continuously monitored measured A common thread between all authentication systems possibility losing forever if attempts manage concur template storage. new trend field avoid utilization original biometrics control processes. possible alternative use cancelable instead. Cancelable can generated through encryption schemes non-invertible transforms. This paper adopts both unified framework ECG signal recognition that used step proposed begins applying transformation fuzzy logic change dynamic range signals. As process nature, it prevents recovery from processed versions, which main target systems. After that, lightweight XOR operation user-specific patterns implemented. Here, high complexity full need large processing burden eliminated. addition stage enhances security traits, allowing hybrid nature merging transforms algorithms. Moreover, an FPGA hardware implementation introduced real ECG-based framework. accompany user allow network when requested. Experimental results show promising performance Area under Receiver Operating Characteristic curve (AROC) 99.5% Equal Error Rate (EER) 0.058%.

Язык: Английский

---

Explainable Enhanced Recurrent Neural Network for lie detection using voice stress analysis

Multimedia Tools and Applications, Год журнала: 2023, Номер 83(11), С. 32277 - 32299

Опубликована: Сен. 20, 2023

Abstract Lie detection is a crucial aspect of human interactions that affects everyone in their daily lives. Individuals often rely on various cues, such as verbal and nonverbal communication, particularly facial expressions, to determine if someone truthful. While automated lie systems can assist identifying these current approaches are limited due lack suitable datasets for testing performance real-world scenarios. Despite ongoing research efforts develop effective reliable methods, this remains work progress. The polygraph, voice stress analysis, pupil dilation analysis some the methods currently used task. In study, we propose new algorithm based an Enhanced Recurrent Neural Network (ERNN) with Explainable AI capabilities. ERNN, long short-term memory (LSTM) architecture, was optimized using fuzzy logic hyperparameters. LSTM model then created trained dataset audio recordings from interviews randomly selected group. proposed ERNN achieved accuracy 97.3%, which statistically significant problem analysis. These results suggest it possible detect patterns voices individuals experiencing explainable manner.

Язык: Английский

---

SleepSmart: an IoT-enabled continual learning algorithm for intelligent sleep enhancement

Neural Computing and Applications, Год журнала: 2023, Номер 36(8), С. 4293 - 4309

Опубликована: Дек. 11, 2023

Abstract Sleep is an essential physiological process that crucial for human health and well-being. However, with the rise of technology increasing work demands, people are experiencing more disrupted sleep patterns. Poor quality quantity can lead to a wide range negative outcomes, including obesity, diabetes, cardiovascular disease. This research paper proposes smart sleeping enhancement system, named SleepSmart, based on Internet Things (IoT) continual learning using bio-signals. The proposed system utilizes wearable biosensors collect data during sleep, which then processed analyzed by IoT platform provide personalized recommendations optimization. Continual techniques employed improve accuracy system's over time. A pilot study subjects was conducted evaluate performance, results show SleepSmart significantly reduce disturbance. has potential practical solution sleep-related issues enhance overall With prevalence problems, be effective tool individuals monitor their quality.

Язык: Английский

---

The effect of consanguineous marriage on reading disability based on deep neural networks

Multimedia Tools and Applications, Год журнала: 2023, Номер 83(17), С. 51787 - 51807

Опубликована: Ноя. 15, 2023

Abstract For knowledge acquisition and social engagement, reading comprehension is essential. However, 20% or so of younger students have trouble with it. In order to predict the effects consanguineous marriage on handicap customize adaptive learning experiences, study proposes an Intelligent Adaptive Learning Prediction Framework (IALPF). This framework proposed as a transformative solution that smoothly combines cutting-edge AI approaches. IALPF provides precise predictions individualized pathways by utilizing extensive cognitive profiling, data gathering, hybrid neural network design. It includes early warning systems, flexible content distribution, ongoing development based active feedback loops. The represents significant change in education has wide-ranging effects. We evaluated skills among 770 included two experimental groups, control group, 22 pupils from first-cousin marriages 21 children unrelated parents, respectively. Tests were given for word identification comprehension, other things. findings showed first cousin parents had higher chance difficulties than those families. outstanding performance IALPF, which outperformed conventional techniques like Back Propagation (BP) General Regression Neural Network (GRNN), was further supported empirical evaluation. demonstrates IALPF's success reinventing personalized predictive analysis, strengthening its potential improve variety scenarios. seamless integration methods into forecasts effect handicap, innovation. To set it apart approaches, this special integrates profile, information networks accurate predictions. analysis revolutionary demonstrating improved accuracy when compared (GRNN).

Язык: Английский

---

Securing Internet-of-Medical-Things networks using cancellable ECG recognition

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Май 13, 2024

Abstract Reinforcement of the Internet Medical Things (IoMT) network security has become extremely significant as these networks enable both patients and healthcare providers to communicate with each other by exchanging medical signals, data, vital reports in a safe way. To ensure transmission sensitive information, robust secure access mechanisms are paramount. Vulnerabilities networks, particularly at points, could expose risks. Among possible measures, biometric authentication is becoming more feasible choice, focus on leveraging regularly-monitored biomedical signals like Electrocardiogram (ECG) due their unique characteristics. A notable challenge within all systems risk losing original traits, if hackers successfully compromise template storage space. Current research endorses replacement biometrics used control cancellable templates. These produced using encryption or non-invertible transformation, which improves enabling templates be changed case an unwanted detected. This study presents comprehensive framework for ECG-based recognition may accessing IoMT networks. An innovative methodology introduced through modification ECG blind signal separation lightweight encryption. The basic idea here depends assumption that auxiliary audio same person subjected algorithm, algorithm will yield two uncorrelated components minimization correlation cost function. Hence, obtained outputs from distorted versions well signals. can treated stage Security enhancement achieved utilization based user-specific pattern XOR operation, thereby reducing processing burden associated conventional methods. proposed efficacy demonstrated its application ECG-ID MIT-BIH datasets, yielding promising results. experimental evaluation reveals Equal Error Rate (EER) 0.134 dataset 0.4 dataset, alongside exceptionally large Area under Receiver Operating Characteristic curve (AROC) 99.96% datasets. results underscore potential securing biometrics, offering hybrid model combines strengths transformations

Язык: Английский

---

Monitoring Physiological State of Drivers Using In-Vehicle Sensing of Non-Invasive Signal

Civil Engineering Journal, Год журнала: 2024, Номер 10(4), С. 1221 - 1231

Опубликована: Апрель 1, 2024

Eighty percent of traffic accidents are caused by human error, called hypo vigilance, stemming from drowsiness, stress, or distraction while driving. This poses a significant threat to road safety. An electrocardiogram (ECG) is often used monitor drivers' health. Thus, enhancing vehicles with Internet Things (IoT) sensors and local analytical databases becomes crucial for real-time detection transmission relevant health data avoid things that compromise study introduces cost-effective in-vehicle ECG sensing prototype using an AD8232 sensor integrated Arduino Uno Wi-Fi module placed on the steering wheel driver's heart signal Short-term rate variability (HRV) features were computed through Python acquired data, supervised machine learning techniques such as AdaBoost, Random Forest, Naïve Bayes, Support Vector Machine (SVM) classified into normal abnormal classes. Naive Bayes exhibited highest accuracy (90.91%) F1 score (85.71%), surpassing Forest's lower (63.64%) (50.00%). These findings indicate prototype's potential valuable tool ensuring safe efficient driving, proposing integration standard vehicle safety systems enhanced Doi: 10.28991/CEJ-2024-010-04-014 Full Text: PDF

Язык: Английский

---

Prediction of Dangerous Driving Behaviour Based on Vehicle Motion

Procedia Computer Science, Год журнала: 2024, Номер 235, С. 1125 - 1134

Опубликована: Янв. 1, 2024

The integration of deep learning, computer vision, and advanced algorithms has ushered in a transformative era the prediction human driving behavior, consequently revolutionizing road safety. This paper focuses on an innovative convergence technology that addresses critical issues like driver fatigue distracted by automatically identifying categorizing diverse behaviors. Neural network architectures, such as VGG16, AlexNet, ResNet are described this have propelled accuracy behavior classification to remarkable levels. However, quest for safer roads remains ongoing, with promising avenues lying ahead. First foremost, creation extensive, diverse, meticulously annotated datasets is paramount. These serve bedrock upon which future models can be trained, enhancing their robustness generalizability across spectrum scenarios. Real-time represent another pivotal frontier. hold potential provide timely interventions support systems drivers, thus preventing accidents proactively. exploration hybrid techniques amalgamate strengths various neural architectures presents exciting avenue, further push boundaries accuracy. Furthermore, also discusses fusion multi-modal data, encompassing sensor data from IoT smartphone devices, holds immense promise. holistic approach promises more comprehensive understanding integrating sources, ultimately contributing environments.In research paper, we explore these cutting-edge developments learning emphasizing technical novelty innovation. Through interdisciplinary approach, envision where synergy technology, leads substantial reduction improved

Язык: Английский

---