Search

Abstract Anticipating human decisions while performing complex tasks remains a formidable challenge. This study proposes a multimodal machine-learning approach that leverages image features and electroencephalography (EEG) data to predict human response correctness in a demanding visual searching task. Notably, we extract a novel set of image features pertaining to object relationships using the Segment Anything Model (SAM), which enhances prediction accuracy compared to traditional features. Additionally, our approach effectively utilizes a combination of EEG signals and image features to streamline the feature set required for the Random Forest Classifier (RFC) while maintaining high accuracy. The findings of this research hold substantial potential for developing advanced fault alert systems, particularly in critical decision-making environments such as the medical and defence sectors.

Smoothing filters are widely used in EEG signal processing for noise removal while preserving signals’ features. Inspired by our recent work on Upscale and Downscale Representation (UDR), this paper proposes a cascade arrangement of some effective image-processing techniques for signal filtering in the image domain. The UDR concept is to visualize EEG signals at an appropriate line width and convert it to a binary image. The smoothing process is then conducted by skeletonizing the signal object to a unit width and projecting it back to the time domain. Two successive UDRs could result in a better-smoothing performance, but their binary image conversion should be restricted. The process is computationally ineffective, especially at higher line width values. Cascaded Thinning UDR (CTUDR) is proposed, exploiting morphological operations to perform a two-stage upscale and downscale within one binary image representation. CTUDR is verified on a signal smoothing and classification task and compared with conventional techniques, such as the Moving Average, the Binomial, the Median, and the Savitzky Golay filters. Simulated EEG data with added white Gaussian noise is employed in the former, while cognitive conflict data obtained from a 3D object selection task is utilized in the latter. CTUDR outperforms its counterparts, scoring the best fitting error and correlation coefficient in signal smoothing while achieving the highest gain in Accuracy (0.7640%) and F-measure (0.7607%) when used as a smoothing filter for training data of EEGNet.

Decoding natural language from noninvasive brain signals has been an exciting topic with the potential to expand the applications of brain-computer interface (BCI) systems. However, current methods face limitations in decoding sentences from electroencephalography (EEG) signals. Improving decoding performance requires the development of a more effective encoder for the EEG modality. Nonetheless, learning generalizable EEG representations remains a challenge due to the relatively small scale of existing EEG datasets. In this paper, we propose enhancing the EEG encoder to improve subsequent decoding performance. Specifically, we introduce the discrete Conformer encoder (D-Conformer) to transform EEG signals into discrete representations and bootstrap the learning process by imposing EEG-language alignment from the early training stage. The D-Conformer captures both local and global patterns from EEG signals and discretizes the EEG representation, making the representation more resilient to variations, while early-stage EEG-language alignment mitigates the limitations of small EEG datasets and facilitates the learning of the semantic representations from EEG signals. These enhancements result in improved EEG representations and decoding performance. We conducted extensive experiments and ablation studies to thoroughly evaluate the proposed method. Utilizing the D-Conformer encoder and bootstrapping training strategy, our approach demonstrates superior decoding performance across various tasks, including word-level, sentence-level, and sentiment-level decoding from EEG signals. Specifically, in word-level classification, we show that our encoding method produces more distinctive representations and higher classification performance compared to the EEG encoders from existing methods. At the sentence level, our model outperformed the baseline by 5.45%, achieving a BLEU-1 score of 42.31%. Furthermore, in sentiment classification, our model exceeded the baseline by 14%, achieving a sentiment classification accuracy of 69.3%.

A significant issue for traffic safety has been drowsy driving for decades. A number of studies have investigated the effects of acute fatigue on spectral power; and recent research has revealed that drowsy driving is associated with a variety of brain connections in a specific cortico-cortical pathway. In spite of this, it is still unclear how different brain regions are connected in drowsy driving at different levels of daily fatigue. This study identified the brain connectivity-behavior relationship among three different daily fatigue levels (low-, median- and high-fatigue) with the EEG data transfer entropy. According to the results, only low- and medium-fatigue groups demonstrated an inverted U-shaped change in connectivity from high performance to poor behavioral performance. In addition, from low- to high-fatigue groups, connectivity magnitude decreased in the frontal region and increased in the occipital region. These study results suggest that brain connectivity and driving behavior would be affected by different levels of daily fatigue.

Drowsy driving is one of the primary causes of driving fatalities. Electroencephalography (EEG), a method for detecting drowsiness directly from brain activity, has been widely used for detecting driver drowsiness in real-time. Recent studies have revealed the great potential of using brain connectivity graphs constructed based on EEG data for drowsy state predictions. However, traditional brain connectivity networks are irrelevant to the downstream prediction tasks. This article proposes a connectivity-aware graph neural network (CAGNN) using a self-attention mechanism that can generate task-relevant connectivity networks via end-to-end training. Our method achieved an accuracy of 72.6% and outperformed other convolutional neural networks (CNNs) and graph generation methods based on a drowsy driving dataset. In addition, we introduced a squeeze-and-excitation (SE) block to capture important features and demonstrated that the SE attention score can reveal the most important feature band. We compared our generated connectivity graphs in the drowsy and alert states and found drowsiness connectivity patterns, including significantly reduced occipital connectivity and interregional connectivity. Additionally, we performed a post hoc interpretability analysis and found that our method could identify drowsiness features such as alpha spindles. Our code is available online at https://github.com/ALEX95GOGO/CAGNN.

Stress is a prevalent bodily response universally experienced and significantly affects a person's mental and cognitive state. The P300 response is a commonly observed brain behaviour that provides insight into a person's cognitive state. Previous works have documented the effects of stress on the P300 behaviour; however, only a few have explored the performance in a mobile and naturalistic experimental setup. Our study examined the effects of stress on the human brain's P300 behaviour through a height exposure experiment that incorporates complex visual, vestibular, and proprioceptive stimuli. A more complex sensory environment could produce translatable findings toward real-world behaviour and benefit emerging technologies such as brain-computer interfaces. Seventeen participants experienced our experiment that elicited the stress response through physical and virtual height exposure. We found two unique groups within our participants that exhibited contrasting behavioural performance and P300 target reaction response when exposed to stressors (from walking at heights). One group performed worse when exposed to heights and exhibited a significant decrease in parietal P300 peak amplitude and increased beta and gamma power. On the other hand, the group less affected by stress exhibited a change in their N170 peak amplitude and alpha/mu rhythm desynchronisation. The findings of our study suggest that a more individualised approach to assessing a person's behaviour performance under stress can aid in understanding P300 performance when experiencing stress.

Object recognition and object identification are multifaceted cognitive operations that require various brain regions to synthesize and process information. Prior research has evidenced the activity of both visual and temporal cortices during these tasks. Notwithstanding their similarities, object recognition and identification are recognized as separate brain functions. Drawing from the two-stream hypothesis, our investigation aims to understand whether the channels within the ventral and dorsal streams contain pertinent information for effective model learning regarding object recognition and identification tasks. By utilizing the data we collected during the object recognition and identification experiment, we scrutinized EEGNet models, trained using channels that replicate the two-stream hypothesis pathways, against a model trained using all available channels. The outcomes reveal that the model trained solely using the temporal region delivered a high accuracy level in classifying four distinct object categories. Specifically, the object recognition and object identification models achieved an accuracy of 89% and 85%, respectively. By incorporating the channels that mimic the ventral stream, the model’s accuracy was further improved, with the object recognition model and object identification model achieving an accuracy of 95% and 94%, respectively. Furthermore, the Grad-CAM result of the trained models revealed a significant contribution from the ventral and dorsal stream channels toward the training of the EEGNet model. The aim of our study is to pinpoint the optimal channel configuration that provides a swift and accurate brain-computer interface system for object recognition and identification.

Identifying meaningful brain activities is critical in brain-computer interface (BCI) applications. Recently, an increasing number of neural network approaches have been proposed to recognize EEG signals. However, these approaches depend heavily on using complex network structures to improve the performance of EEG recognition and suffer from the deficit of training data. Inspired by the waveform characteristics and processing methods shared between EEG and speech signals, we propose Speech2EEG, a novel EEG recognition method that leverages pretrained speech features to improve the accuracy of EEG recognition. Specifically, a pretrained speech processing model is adapted to the EEG domain to extract multichannel temporal embeddings. Then, several aggregation methods, including the weighted average, channelwise aggregation, and channel-and-depthwise aggregation, are implemented to exploit and integrate the multichannel temporal embeddings. Finally, a classification network is used to predict EEG categories based on the integrated features. Our work is the first to explore the use of pretrained speech models for EEG signal analysis as well as the effective ways to integrate the multichannel temporal embeddings from the EEG signal. Extensive experimental results suggest that the proposed Speech2EEG method achieves state-of-the-art performance on two challenging motor imagery (MI) datasets, the BCI IV-2a and BCI IV-2b datasets, with accuracies of ${89}.{5}\%$ and ${84}.{07}\%$ , respectively. Visualization analysis of the multichannel temporal embeddings show that the Speech2EEG architecture can capture useful patterns related to MI categories, which can provide a novel solution for subsequent research under the constraints of a limited dataset scale.

Research and development of electroencephalogram (EEG) based brain-computer interfaces (BCIs) have advanced rapidly, partly due to deeper understanding of the brain and wide adoption of sophisticated machine learning approaches for decoding the EEG signals. However, recent studies have shown that machine learning algorithms are vulnerable to adversarial attacks. This paper proposes to use narrow period pulse for poisoning attack of EEG-based BCIs, which makes adversarial attacks much easier to implement. One can create dangerous backdoors in the machine learning model by injecting poisoning samples into the training set. Test samples with the backdoor key will then be classified into the target class specified by the attacker. What most distinguishes our approach from previous ones is that the backdoor key does not need to be synchronized with the EEG trials, making it very easy to implement. The effectiveness and robustness of the backdoor attack approach is demonstrated, highlighting a critical security concern for EEG-based BCIs and calling for urgent attention to address it.

Abstract Although beacon- and map-based spatial strategies are the default strategies for navigation activities, today’s navigational aids mostly follow a beacon-based design where one is provided with turn-by-turn instructions. Recent research, however, shows that our reliance on these navigational aids is causing a decline in our spatial skills. We are processing less of our surrounding environment and relying too heavily on the instructions given. To reverse this decline, we need to engage more in map-based learning, which encourages the user to process and integrate spatial knowledge into a cognitive map built to benefit flexible and independent spatial navigation behaviour. In an attempt to curb our loss of skills, we proposed a navigation assistant to support map-based learning during active navigation. Called the virtual global landmark (VGL) system, this augmented reality (AR) system is based on the kinds of techniques used in traditional orienteering. Specifically, a notable landmark is always present in the user’s sight, allowing the user to continuously compute where they are in relation to that specific location. The efficacy of the unit as a navigational aid was tested in an experiment with 27 students from the University of Technology Sydney via a comparison of brain dynamics and behaviour. From an analysis of behaviour and event-related spectral perturbation, we found that participants were encouraged to process more spatial information with a map-based strategy where a silhouette of the compass-like landmark was perpetually in view. As a result of this technique, they consistently navigated with greater efficiency and better accuracy.

Abstract To reduce the decline of spatial cognitive skills caused by the increasing use of automated GPS navigation, the virtual global landmark (VGL) system is proposed to help people naturally improve their sense of direction. Designed to accompany a heads-up navigation system, VGL system constantly displays silhouette of global landmarks in the navigator’s vision as a notable frame of reference. This study exams how VGL system impacts incidental spatial learning, i.e., subconscious spatial knowledge acquisition. We asked 55 participants to explore a virtual environment and then draw a map of what they had explored while capturing electroencephalogram (EEG) signals and eye activity. The results suggest that, with the VGL system, participants paid more attention during exploration and performed significantly better at the map drawing task—a result that indicates substantially improved incidental spatial learning. This finding might kickstart a redesigning navigation aids, to teach users to learn a route rather than simply showing them the way.

Modern work environments have extensive interactions with technology and greater cognitive complexity of the tasks, which results in human operators experiencing increased mental workload. Air traffic control operators routinely work in such complex environments, and we designed tracking and collision prediction tasks to emulate their elementary tasks. The physiological response to the workload variations in these tasks was elucidated to untangle the impact of workload variations experienced by operators. Electroencephalogram (EEG), eye activity, and heart rate variability (HRV) data were recorded from 24 participants performing tracking and collision prediction tasks with three levels of difficulty. Our findings indicate that variations in task load in both these tasks are sensitively reflected in EEG, eye activity and HRV data. Multiple regression results also show that operators’ performance in both tasks can be predicted using the corresponding EEG, eye activity and HRV data. The results also demonstrate that the brain dynamics during each of these tasks can be estimated from the corresponding eye activity, HRV and performance data. Furthermore, the markedly distinct neurometrics of workload variations in the tracking and collision prediction tasks indicate that neurometrics can provide insights on the type of mental workload. These findings have applicability to the design of future mental workload adaptive systems that integrate neurometrics in deciding not just “when” but also “what” to adapt. Our study provides compelling evidence in the viability of developing intelligent closed-loop mental workload adaptive systems that ensure efficiency and safety in complex work environments.

Correct detection of peaks in electroencephalogram (EEG) signals is of essence due to the significant correlation of those potentials with cognitive performance and disorders. This paper proposes a novel and non-parametric approach to detect prediction error negativity (PEN) in cognitive conflict processing. The PEN candidates are first located from the input signal via an adaptation of a recent effective method for local maxima extraction, processed in a multi-scale manner. The found candidates are then fused and ranked based on their shape and location-based features. False positives caused by candidates’ magnitude are eliminated by rotating the sorted candidate list where the one with the second-best ranking score will be identified as PEN. The EEG data collected from a 3D object selection task have been used to verify the efficacy of the proposed approach. Compared with the state-of-the-art peak detection techniques, the proposed method shows an improvement of at least 2.67% in accuracy and 6.27% in sensitivity while requires only about 4 ms to process an epoch. The accuracy and computational efficiency of the proposed technique in the detection of PEN in cognitive conflict processing would lead to promising applications in performance improvement of brain-computer interfaces (BCIs).

Analyzing the effects landmarks have on spatial learning is an active area of research in the study of human navigation processes and one that is key to understanding the links between human brain dynamics, landmark encoding, and spatial learning outcomes. This article presents a study on whether electroencephalography (EEG) signals related to virtual global landmarks combined with deep learning can be used to predict the accuracy and efficacy of spatial learning. Virtual global landmarks are silhouettes of actual landmarks projected into the navigator’s vision via a heads-up display. They serve as a notable frame of reference in addition to the local landmarks we all typically use for route navigation. From a mobile virtual reality scenario involving 55 participants, the results of the study suggest that the EEG data associated with those who were exposed to global landmarks shows a visibly better capacity for predicting the quality of spatial learning levels than those who were not. As such, the EEG features associated with processing VGLs have a greater functional relation to the quality of spatial learning. This finding opens up a future direction of enquiry into landmark encoding and navigational ability. It may also provide a potential avenue for the early diagnosis of Alzheimer’s disease.

Motor imagery-based brain-computer interface (MI-BCI) currently represents a new trend in rehabilitation. However, individual differences in the responsive frequency bands and a poor understanding of the communication between the ipsilesional motor areas and other regions limit the use of MI-BCI therapy. Objective: Bimanual training has recently attracted attention as it achieves better outcomes as compared to repetitive one-handed training. This study compared the effects of three MI tasks with different visual feedback. Methods: Fourteen healthy subjects performed single hand motor imagery tasks while watching single static hand (traditional MI), single hand with rotation movement (rmMI), and bimanual coordination with a hand pedal exerciser (bcMI). Functional connectivity is estimated by Transfer Entropy (TE) analysis for brain information flow. Results: Brain connectivity of conducting three MI tasks showed that the bcMI demonstrated increased communications from the parietal to the bilateral prefrontal areas and increased contralateral connections between motor-related zones and spatial processing regions. Discussion/Conclusion: The results revealed bimanual coordination operation events increased spatial information and motor planning under the motor imagery task. And the proposed bimanual coordination MI-BCI (bcMI-BCI) can also achieve the effect of traditional motor imagery tasks and promotes more effective connections with different brain regions to better integrate motor-cortex functions for aiding the development of more effective MI-BCI therapy. Clinical and Translational Impact Statement The proposed bcMI-BCI provides more effective connections with different brain areas and integrates motor-cortex functions to promote motor imagery rehabilitation for patients’ impairment.

Bio-signals pattern recognition systems can be impacted by several factors with a potential to limit their associated performance and clinical translation. Among these factors, selecting the optimum feature extraction method, that can effectively exploit the interaction between the temporal and spatial information, is the most prominent. Despite the potential of deep learning (DL) models for extracting temporal, spatial, or temporal-spatial information, they are typically restricted by their need for a large amount of training data. The deep wavelet scattering transform (WST) is a relatively recent advancement within the DL literature to replace expensive convolution neural networks models with computationally less demanding methods. However, while some studies have used WST to extract features from biological signals, it has not been investigated before for electromyogram (EMG) and electroencephalogram (EEG) signals feature extraction. To investigate the hypothesis of the usefulness of WST for processing EMG and EEG signals, this study used a tremor dataset collected by the authors from people with tremor disorders. Specifically, the proposed work achieved three goals: (a) study the performance of extracting features from low-density EMG signals (8 channels), using the WST approach, (b) study the effect of extracting the features from high-density EEG signals (33 channels), using WST and study its robustness against changing the spatial and temporal aspects of classification accuracy, and (c) classify tremor severity using the WST method and compare the results with other well-known feature extraction approaches. The classification error rates were significantly reduced (maximum of nearly 12%) compared with other feature sets.

Wearable smart glasses are an emerging technology gaining popularity in the assistive technologies industry. Smart glasses aids typically leverage computer vision and other sensory information to translate the wearer's surrounding into computer-synthesized speech. In this work, we explored the potential of a new technique known as "acoustic touch" to provide a wearable spatial audio solution for assisting people who are blind in finding objects. In contrast to traditional systems, this technique uses smart glasses to sonify objects into distinct sound auditory icons when the object enters the device's field of view. We developed a wearable Foveated Audio Device to study the efficacy and usability of using acoustic touch to search, memorize, and reach items. Our evaluation study involved 14 participants, 7 blind or low-visioned and 7 blindfolded sighted (as a control group) participants. We compared the wearable device to two idealized conditions, a verbal clock face description and a sequential audio presentation through external speakers. We found that the wearable device can effectively aid the recognition and reaching of an object. We also observed that the device does not significantly increase the user's cognitive workload. These promising results suggest that acoustic touch can provide a wearable and effective method of sensory augmentation.

Abstract Spatial navigation is a complex cognitive process based on multiple senses that are integrated and processed by a wide network of brain areas. Previous studies have revealed the retrosplenial complex (RSC) to be modulated in a task-related manner during navigation. However, these studies restricted participants’ movement to stationary setups, which might have impacted heading computations due to the absence of vestibular and proprioceptive inputs. Here, we present evidence of human RSC theta oscillation (4–8 Hz) in an active spatial navigation task where participants actively ambulated from one location to several other points while the position of a landmark and the starting location were updated. The results revealed theta power in the RSC to be pronounced during heading changes but not during translational movements, indicating that physical rotations induce human RSC theta activity. This finding provides a potential evidence of head-direction computation in RSC in healthy humans during active spatial navigation.

Objective: Mental stress is a major problem in our society and has become an area of interest for many psychiatric researchers. One primary research focus area is the identification of bio-markers that not only identify stress but also predict the conditions (or tasks) that cause stress. Electroencephalograms (EEGs) have been used for a long time to study and identify bio-markers. While these bio-markers have successfully predicted stress in EEG studies for binary conditions, their performance is suboptimal for multiple conditions of stress. Methods: To overcome this challenge, we propose using latent based representations of the bio-markers, which have been shown to significantly improve EEG performance compared to traditional bio-markers alone. We evaluated three commonly used EEG based bio-markers for stress, the brain load index (BLI), the spectral power values of EEG frequency bands (alpha, beta and theta), and the relative gamma (RG), with their respective latent representations using four commonly used classifiers. Results: The results show that spectral power value based bio-markers had a high performance with an accuracy of 83%, while the respective latent representations had an accuracy of 91%.

Brain-computer interfaces (BCI) relying on electroencephalography (EEG) based neuroimaging mode has shown prospects for real-world usage due to its portability and optional selectivity of fewer channels for compactness. However, noise and artifacts often limit the capacity of BCI systems especially for event-related potentials such as P300 and error-related negativity (ERN), whose biomarkers are present in short time segments at the time-series level. Contrary to EEG, invasive recording is less prone to noise but requires a tedious surgical procedure. But EEG signal is the result of aggregation of neuronal spiking information underneath the scalp surface and transforming the relevant BCI task's EEG signal to spike representation could potentially help improve the BCI performance. In this study, we designed an approach using a spiking neural network (SNN) which is trained using surrogate-gradient descent to generate task-related multi-channel EEG template signals of all classes. The trained model is in turn leveraged to obtain the latent spike representation for each EEG sample. Comparing the classification performance of EEG signal and its spike-representation, the proposed approach enhanced the performance of ERN dataset from 79.22 to 82.27% with naive bayes and for P300 dataset, the accuracy was improved from 67.73 to 69.87% using xGboost. In addition, principal component analysis and correlation metrics were evaluated on both EEG signals and their spike-representation to identify the reason for such improvement.

Freezing of gait (FOG) severely incapacitates the mobility of patients with advanced Parkinson’s disease (PD). An accurate prediction of the onset of FOG could improve the quality of life for PD patients. However, it is imperative to distinguish the possibility of the onset of FOG from that of voluntary stopping. Our previous work demonstrated the neurological differences between the transition to FOG and voluntary stopping using electroencephalogram (EEG) signals. We employed a timed up-and-go (TUG) task to elicit FOG in PD patients. Some of these TUG tasks had an additional voluntary stopping component, where participants stopped walking based on verbal instruction to “stop”. The performance of the convolutional neural network (CNN) in identifying the transition to FOG from normal walking and the transition to voluntary stopping was explored. To the best of our knowledge, this work is the first study to propose a deep learning method to distinguish the transition to FOG from the transition to voluntary stop in PD patients. The models, trained on the EEG data from 17 PD patients who manifested FOG episodes, considering a short two-second transition window for FOG occurrence or voluntary stopping, achieved close to 75% classification accuracy in distinguishing transition to FOG from the transition to voluntary stopping or normal walking. Our results represent an important step toward advanced EEG-based cueing systems for smart FOG intervention, excluding the potential confounding of voluntary stopping.

The K-means algorithm is a widely used clustering algorithm that offers simplicity and efficiency. However, the traditional K-means algorithm uses a random method to determine the initial cluster centers, which make clustering results prone to local optima and then result in worse clustering performance. In this research, we propose an adaptive initialization method for the K-means algorithm (AIMK) which can adapt to the various characteristics in different datasets and obtain better clustering performance with stable results. For larger or higher-dimensional datasets, we even leverage random sampling in AIMK (name as AIMK-RS) to reduce the time complexity. 22 real-world datasets were applied for performance comparisons. The experimental results show AIMK and AIMK-RS outperform the current initialization methods and several well-known clustering algorithms. Specifically, AIMK-RS can significantly reduce the time complexity to O (n). Moreover, we exploit AIMK to initialize K-medoids and spectral clustering, and better performance is also explored. The above results demonstrate superior performance and good scalability by AIMK or AIMK-RS. In the future, we would like to apply AIMK to more partition-based clustering algorithms to solve real-life practical problems.

The modelling of trust values on agents is broadly considered fundamental for decision-making in human-autonomous teaming (HAT) systems. Compared to the evaluation of trust values for robotic agents, estimating human trust is more challenging due to trust miscalibration issues, including undertrust and overtrust problems. From a subjective perception, human trust could be altered along with dynamic human cognitive states, which makes trust values hard to calibrate properly. Thus, in an attempt to capture the dynamics of human trust, the present study evaluated the dynamic nature of trust for human agents through real-time multievidence measures, including human states of attention, stress and perception abilities. The proposed multievidence human trust model applied an adaptive fusion method based on fuzzy reinforcement learning to fuse multievidence from eye trackers, heart rate monitors and human awareness. In addition, fuzzy reinforcement learning was applied to generate rewards via a fuzzy logic inference process that has tolerance for uncertainty in human physiological signals. The results of robot simulation suggest that the proposed trust model can generate reliable human trust values based on real-time cognitive states in the process of ongoing tasks. Moreover, the human-autonomous team with the proposed trust model improved the system efficiency by over 50% compared to the team with only autonomous agents. These results may demonstrate that the proposed model could provide insight into the real-time adaptation of HAT systems based on human states and, thus, might help develop new ways to enhance future HAT systems better.

Organised attacks on a computer system to test existing defences, i.e., penetration testing, have been used extensively to evaluate network security. However, penetration testing is a time-consuming process. Additionally, establishing a strategy that resembles a real cyber-attack typically requires in-depth knowledge of the cybersecurity domain. This paper presents a novel architecture, named deep cascaded reinforcement learning agents, or CRLA, that addresses large discrete action spaces in an autonomous penetration testing simulator, where the number of actions exponentially increases with the complexity of the designed cybersecurity network. Employing an algebraic action decomposition strategy, CRLA is shown to find the optimal attack policy in scenarios with large action spaces faster and more stably than a conventional deep Q-learning agent, which is commonly used as a method for applying artificial intelligence to autonomous penetration testing.

Electrooculography (EOG) is considered as the most stable physiological signal in the development of human-computer interface (HCI) for detecting eye-movement variations. EOG signal classification has gained more traction in recent years to overcome physical inconvenience in paralyzed patients. In this paper, a robust classification technique, such as eight directional movements is investigated by introducing a concept of buffer along with a variation of the slope to avoid misclassification effects in EOG signals. Blinking detection becomes complicated when the magnitude of the signals are considered. Hence, a correction technique is introduced to avoid misclassification for oblique eye movements. Meanwhile, a case study has been considered to apply these correction techniques to HCI baseball game to learn eye-movements.

This paper discusses a novel approach to an EEG (electroencephalogram)-based driver distraction classification by using brain connectivity estimators as features. Ten healthy volunteers with more than one year of driving experience and an average age of 24.3 participated in a virtual reality environment with two conditions, a simple math problem-solving task and a lane-keeping task to mimic the distracted driving task and a non-distracted driving task, respectively. Independent component analysis (ICA) was conducted on the selected epochs of six selected components relevant to the frontal, central, parietal, occipital, left motor, and right motor areas. Granger–Geweke causality (GGC), directed transfer function (DTF), partial directed coherence (PDC), and generalized partial directed coherence (GPDC) brain connectivity estimators were used to calculate the connectivity matrixes. These connectivity matrixes were used as features to train the support vector machine (SVM) with the radial basis function (RBF) and classify the distracted and non-distracted driving tasks. GGC, DTF, PDC, and GPDC connectivity estimators yielded the classification accuracies of 82.27%, 70.02%, 86.19%, and 80.95%, respectively. Further analysis of the PDC connectivity estimator was conducted to determine the best window to differentiate between the distracted and non-distracted driving tasks. This study suggests that the PDC connectivity estimator can yield better classification accuracy for driver distractions.

With the advent of advanced machine learning methods, the performance of brain–computer interfaces (BCIs) has improved unprecedentedly. However, electroencephalography (EEG), a commonly used brain imaging method for BCI, is characterized by a tedious experimental setup, frequent data loss due to artifacts, and is time consuming for bulk trial recordings to take advantage of the capabilities of deep learning classifiers. Some studies have tried to address this issue by generating artificial EEG signals. However, a few of these methods are limited in retaining the prominent features or biomarker of the signal. And, other deep learning-based generative methods require a huge number of samples for training, and a majority of these models can handle data augmentation of one category or class of data at any training session. Therefore, there exists a necessity for a generative model that can generate synthetic EEG samples with as few available trials as possible and generate multi-class while retaining the biomarker of the signal. Since EEG signal represents an accumulation of action potentials from neuronal populations beneath the scalp surface and as spiking neural network (SNN), a biologically closer artificial neural network, communicates via spiking behavior, we propose an SNN-based approach using surrogate-gradient descent learning to reconstruct and generate multi-class artificial EEG signals from just a few original samples. The network was employed for augmenting motor imagery (MI) and steady-state visually evoked potential (SSVEP) data. These artificial data are further validated through classification and correlation metrics to assess its resemblance with original data and in-turn enhanced the MI classification performance.

Detecting and correcting incorrect body movements is an essential part of everyday interaction with one's environment. The human brain provides a monitoring system that constantly controls and adjusts our actions according to our surroundings. However, when our brain's predictions about a planned action do not match the sensory inputs resulting from that action, cognitive conflict occurs. Much is known about cognitive conflict in 1D/2D environments; however, less is known about the role of movement characteristics associated with cognitive conflict in 3D environment. Hence, we devised an object selection task in a virtual reality (VR) environment to test how the velocity of hand movements impacts human brain responses. From a series of analyses of EEG recordings synchronized with motion capture, we found that the velocity of the participants’ hand movements modulated the brain's response to proprioceptive feedback during the task and induced a prediction error negativity (PEN). Additionally, the PEN originates in the anterior cingulate cortex and is itself modulated by the ballistic phase of the hand's movement. These findings suggest that velocity is an essential component of integrating hand movements with visual and proprioceptive information during interactions with real and virtual objects.

Many studies have reported that exercise can influence cognitive performance. But advancing our understanding of the interrelations between psychology and physiology in sports neuroscience requires the study of real-time brain dynamics during exercise in the field. Electroencephalography (EEG) is one of the most powerful brain imaging technologies. However, the limited portability and long preparation time of traditional wet-sensor systems largely limits their use to laboratory settings. Wireless dry-sensor systems are emerging with much greater potential for practical application in sports. Hence, in this paper, we use the BR8 wireless dry-sensor EEG system to measure P300 brain dynamics while cycling at various intensities. The preparation time was mostly less than 2 min as BR8 system’s dry sensors were able to attain the required skin-sensor interface impedance, enabling its operation without any skin preparation or application of conductive gel. Ten participants performed four sessions of a 3 min rapid serial visual presentation (RSVP) task while resting and while cycling. These four sessions were pre-CE (RSVP only), low-CE (RSVP in 40–50% of max heart rate), vigorous-CE (RSVP in 71–85% of max heart rate) and post-CE (RSVP only). The recorded brain signals demonstrate that the P300 amplitudes, observed at the Pz channel, for the target and non-target responses were significantly different in all four sessions. The results also show decreased reaction times to the visual attention task during vigorous exercise, enriching our understanding of the ways in which exercise can enhance cognitive performance. Even though only a single channel was evaluated in this study, the quality and reliability of the measurement using these dry sensor-based EEG systems is clearly demonstrated by our results. Further, the smooth implementation of the experiment with a dry system and the success of the data analysis demonstrate that wireless dry EEG devices can open avenues for real-time measurement of cognitive functions in athletes outside the laboratory.

For emergency or intensive-care units (ICUs), patients with unclear consciousness or unstable hemodynamics often require aggressive monitoring by multiple monitors. Complicated pipelines or lines increase the burden on patients and inconvenience for medical personnel. Currently, many commercial devices provide related functionalities. However, most devices measure only one biological signal, which can increase the budget for users and cause difficulty in remote integration. In this study, we develop a wearable device that integrates electrocardiography (ECG), electroencephalography (EEG), and blood oxygen machines for medical applications with the hope that it can be applied in the future. We develop an integrated multiple-biosignal recording system based on a modular design. The developed system monitors and records EEG, ECG, and peripheral oxygen saturation (SpO2) signals for health purposes simultaneously in a single setting. We use a logic level converter to connect the developed EEG module (BR8), ECG module, and SpO2 module to a microcontroller (Arduino). The modular data are then smoothly encoded and decoded through consistent overhead byte stuffing (COBS). This developed system has passed simulation tests and exhibited proper functioning of all modules and subsystems. In the future, the functionalities of the proposed system can be expanded with additional modules to support various emergency or ICU applications.

We present a hierarchical fuzzy logic system for precision coordination of multiple mobile agents such that they achieve simultaneous arrival at their destination positions in a cluttered urban environment. We assume that each agent is equipped with a 2D scanning Lidar to make movement decisions based on local distance and bearing information. Two solution approaches are considered and compared. Both of them are structured around a hierarchical arrangement of control modules to enable synchronization of the agents' arrival times while avoiding collision with obstacles. The proposed control module controls both moving speeds and directions of the robots to achieve the simultaneous target-reaching task. The control system consists of two levels: the lower-level individual navigation control for obstacle avoidance and the higher-level coordination control to ensure the same time of arrival for all robots at their target. The first approach is based on cascading fuzzy logic controllers, and the second approach considers the use of a Long Short-Term Memory recurrent neural network module alongside fuzzy logic controllers. The parameters of all the controllers are optimized using the particle swarm optimization algorithm. To increase the scalability of the proposed control modules, an interpolation method is introduced to determine the velocity scaling factors and the searching directions of the robots. A physics-based simulator, Webots, is used as a training and testing environment for the two learning models to facilitate the deployment of codes to hardware, which will be conducted in the next phase of our research.

Previous studies have shown equivocal results about whether atypical or unusual events, compared with typical ones, facilitate or inhibit memory. We suspect that the indefinite findings could be partly due to the recall task used in these studies, as the participants might have used inference instead of recall in their responses. In the present study, we tested the recognition memory for real (Experiment 1) and fabricated (Experiment 2) advertisements, which could be congruent or incongruent with gender stereotypes. In congruent advertisements, a female endorser presented a traditionally considered feminine product or a male endorser presented a traditionally considered masculine product, whereas the gender-product type matching reversed in incongruent advertisements. The results of both behavioral experiments revealed that the participants’ memory performance for stereotype-incongruent advertisements was higher than for congruent ones. In the event-related potential (ERP) recordings in Experiment 3, larger positive amplitudes were found for stereotype-incongruent advertisements than for congruent advertisements on the left parietal sites, suggesting a deeper encoding process for stereotype-incongruent information than for stereotype-congruent information.

In the above article [1], we detected an error in reporting the 10-fold cross-validation result. The correct 10-fold cross-validation result in Table IV is uploaded in this letter.

Abstract Background In the past decade, fatigue has been regarded as one of the main factors impairing task performance and increasing behavioral lapses during driving, even leading to fatal car crashes. Although previous studies have explored the impact of acute fatigue through electroencephalography (EEG) signals, it is still unclear how different fatigue levels affect brain–behavior relationships. Methods A longitudinal study was performed to investigate the brain dynamics and behavioral changes in individuals under different fatigue levels by a sustained attention task. This study used questionnaires in combination with actigraphy, a noninvasive means of monitoring human physiological activity cycles, to conduct longitudinal assessment and tracking of the objective and subjective fatigue levels of recruited participants. In this study, degrees of effectiveness score (fatigue rating) are divided into three levels (normal, reduced, and high risk) by the SAFTE fatigue model. Results Results showed that those objective and subjective indicators were negatively correlated to behavioral performance. In addition, increased response times were accompanied by increased alpha and theta power in most brain regions, especially the posterior regions. In particular, the theta and alpha power dramatically increased in the high‐fatigue (high‐risk) group. Additionally, the alpha power of the occipital regions showed an inverted U‐shaped change. Conclusion Our results help to explain the inconsistent findings among existing studies, which considered the effects of only acute fatigue on driving performance while ignoring different levels of resident fatigue, and potentially lead to practical and precise biomathematical models to better predict the performance of human operators.

Introduction: We hypothesized that hypertensive patients harbor a higher risk of urinary bladder (UB) cancer. Material and methods: We performed a population-based cohort study on adults using a National Health Insurance Research Database (NHIRD) dataset. Hypertension and comparison non-hypertensive (COMP) groups comprising 39,618 patients each were propensity score-matched by age, sex, index date, and medical comorbidities. The outcome was incident UB cancer validated using procedure codes. We constructed multivariable Cox models to derive adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs). Cumulative incidence was compared using a log-rank test. Results: During a total follow-up duration of 380,525 and 372,020 person-years in the hypertension and COMP groups, 248 and 186 patients developed UB cancer, respectively, representing a 32% increase in the risk (aHR, 1.32; 95% CI, 1.09–1.60). Hypertensive women harbored a significantly increased risk of UB cancer (aHR, 1.55; 95% CI, 1.12–2.13) compared with non-hypertensive women, whereas men with hypertension had a statistically non-significant increased risk (aHR, 1.22; 95% CI, 0.96–1.55). The sensitivity analysis demonstrated that the increased risk was sustained throughout different follow-up durations for the entire cohort; a statistical increase in the risk was also noted among hypertensive men. Conclusion: This nationwide population-based propensity score-matched cohort study supports a positive association between hypertension and subsequent UB cancer development.

Polysomnography (PSG) is considered the gold standard in the diagnosis of obstructive sleep apnea (OSA). The diagnosis of OSA requires an overnight sleep experiment in a laboratory. However, due to limitations in relation to the number of labs and beds available, patients often need to wait a long time before being diagnosed and eventually treated. In addition, the unfamiliar environment and restricted mobility when a patient is being tested with a polysomnogram may disturb their sleep, resulting in an incomplete or corrupted test. Therefore, it is posed that a PSG conducted in the patient's home would be more reliable and convenient. The Internet of Things (IoT) plays a vital role in the e-Health system. In this paper, we implement an IoT-based wireless polysomnography system for sleep monitoring, which utilizes a batterypowered, miniature, wireless, portable, and multipurpose recorder. A Java-based PSG recording program in the personal computer is designed to save several bio-signals and transfer them into the European data format. These PSG records can be used to determine a patient's sleep stages and diagnose OSA. This system is portable, lightweight, and has low power-consumption. To demonstrate the feasibility of the proposed PSG system, a comparison was made between the standard PSG-Alice 5 Diagnostic Sleep System and the proposed system. Several healthy volunteer patients participated in the PSG experiment and were monitored by both the standard PSG-Alice 5 Diagnostic Sleep System and the proposed system simultaneously, under the supervision of specialists at the Sleep Laboratory in Taipei Veteran General Hospital. A comparison of the results of the time-domain waveform and sleep stage of the two systems shows that the proposed system is reliable and can be applied in practice. The proposed system can facilitate the long-term tracing and research of personal sleep monitoring at home.

Task switching is a common method to investigate executive functions such as working memory and attention. This paper investigates the effect of acute stress on brain activity using task switching. Surprisingly few studies have been conducted in this area. There is behavioral and physiological evidence to indicate that acute stress makes the participants more tense which results in a better performance. In this current study, under stressful conditions, the participants gave quick responses with high accuracy. However, unexpected results were found in relation to salivary cortisol. Furthermore, the electroencephalogram results showed that acute stress was pronounced at the frontal and parietal midline cortex, especially on the theta, alpha, and gamma bands. One possible explanation for these results may be that the participants changed their strategy in relation to executive functions during stressful conditions by paying more attention which resulted in a higher working memory capacity which enhanced performance during the task switching.

A centralized infrastructure system carries out existing data analytics and decision-making processes from our current highly virtualized platform of wireless networks and the Internet of Things (IoT) applications. There is a high possibility that these existing methods will encounter more challenges and issues in relation to network dynamics, resulting in a high overhead in the network response time, leading to latency and traffic. In order to avoid these problems in the network and achieve an optimum level of resource utilization, a new paradigm called edge computing (EC) is proposed to pave the way for the evolution of new age applications and services. With the integration of EC, the processing capabilities are pushed to the edge of network devices such as smart phones, sensor nodes, wearables, and on-board units, where data analytics and knowledge generation are performed which removes the necessity for a centralized system. Many IoT applications, such as smart cities, the smart grid, smart traffic lights, and smart vehicles, are rapidly upgrading their applications with EC, significantly improving response time as well as conserving network resources. Irrespective of the fact that EC shifts the workload from a centralized cloud to the edge, the analogy between EC and the cloud pertaining to factors such as resource management and computation optimization are still open to research studies. Hence, this paper aims to validate the efficiency and resourcefulness of EC. We extensively survey the edge systems and present a comparative study of cloud computing systems. After analyzing the different network properties in the system, the results show that EC systems perform better than cloud computing systems. Finally, the research challenges in implementing an EC system and future research directions are discussed.

Automated human action recognition has the potential to play an important role in public security, for example, in relation to the multiview surveillance videos taken in public places, such as train stations or airports. This paper compares three practical, reliable, and generic systems for multiview video-based human action recognition, namely, the nearest neighbor classifier, Gaussian mixture model classifier, and the nearest mean classifier. To describe the different actions performed in different views, view-invariant features are proposed to address multiview action recognition. These features are obtained by extracting the holistic features from different temporal scales which are modeled as points of interest which represent the global spatial-temporal distribution. Experiments and cross-data testing are conducted on the KTH, WEIZMANN, and MuHAVi datasets. The system does not need to be retrained when scenarios are changed which means the trained database can be applied in a wide variety of environments, such as view angle or background changes. The experiment results show that the proposed approach outperforms the existing methods on the KTH and WEIZMANN datasets.

Different rendering styles induce different levels of agency and user behaviors in virtual reality environments. We applied an electroencephalogram-based approach to investigate how the rendering style of the users' hands affects behavioral and cognitive responses. To this end, we introduced prediction errors due to cognitive conflicts during a 3-D object selection task by manipulating the selection distance of the target object. The results showed that, for participants with high behavioral inhibition scores, the amplitude of the negative event-related potential at approximately 50-250 ms correlated with the realism of the virtual hands. Concurring with the uncanny valley theory, these findings suggest that the more realistic the representation of the user's hand is, the more sensitive the user becomes toward subtle errors, such as tracking inaccuracies.

Improving the degree of assistance given by in-car navigation systems is an important issue for the safety of both drivers and passengers. There is a vast body of research that assesses the usability and interfaces of the existing navigation systems but very few investigations study the impact on the brain activity based on navigation-based driving. In this paper, a real-world experiment is designed to acquire the electroencephalography (EEG) and in-car information to analyze the dynamic brain activity while the driver is performing the lane-changing task based on the auditory instructions from an in-car navigation system. The results show that auditory cues can influence the speed and increase the frontal EEG delta and beta power, which is related to motor preparation and decision making during a lane change. However, there were no significant results on the alpha power. A better lane-change assessment can be obtained using specific vehicle information (lateral acceleration and heading angle) with EEG features for future naturalized driving study.

Stress is an inherent part of the normal human experience. Although, for the most part, this stress response is advantageous, chronic, heightened, or inappropriate stress responses can have deleterious effects on the human body. It has been suggested that individuals who experience repeated or prolonged stress exhibit blunted biological stress responses when compared to the general population. Thus, when assessing whether a ubiquitous stress response exists, it is important to stratify based on resting levels in the absence of stress. Research has shown that stress that causes symptomatic responses requires early intervention in order to mitigate possible associated mental health decline and personal risks. Given this, real-time monitoring of stress may provide immediate biofeedback to the individual and allow for early self-intervention. This study aimed to determine if the change in heart rate variability could predict, in two different cohorts, the quality of response to acute stress when exposed to an acute stressor and, in turn, contribute to the development of a physiological algorithm for stress which could be utilized in future smartwatch technologies. This study also aimed to assess whether baseline stress levels may affect the changes seen in heart rate variability at baseline and following stress tasks. A total of 30 student doctor participants and 30 participants from the general population were recruited for the study. The Trier Stress Test was utilized to induce stress, with resting and stress phase ECGs recorded, as well as inter-second heart rate (recorded using a FitBit). Although the present study failed to identify ubiquitous patterns of HRV and HR changes during stress, it did identify novel changes in these parameters between resting and stress states. This study has shown that the utilization of HRV as a measure of stress should be calculated with consideration of resting (baseline) anxiety and stress states in order to ensure an accurate measure of the effects of additive acute stress.

A new online multi-class learning algorithm is proposed with three main characteristics. First, in order to make the feature pool fitter for the pattern pool, the adaptive feature pool is proposed to dynamically combine the three general features, Haar-like, Histogram of Oriented Gradient (HOG), and Local Binary Patterns (LBP). Second, the external model is integrated into the proposed model without re-training to enhance the efficacy of the model. Third, a new multi-class learning and updating mechanism are proposed that help to find unsuitable decisions and adjust them automatically. The performance of the proposed model is validated with multi-class detection and online learning system. The proposed model achieves a better score than other non-deep learning algorithms used in public pedestrian and multi-class databases. The multi-class databases contain data for pedestrians, faces, vehicles, motorcycles, bicycles, and aircraft.

In the assistive research area, human–computer interface (HCI) technology is used to help people with disabilities by conveying their intentions and thoughts to the outside world. Many HCI systems based on eye movement have been proposed to assist people with disabilities. However, due to the complexity of the necessary algorithms and the difficulty of hardware implementation, there are few general-purpose designs that consider practicality and stability in real life. Therefore, to solve these limitations and problems, an HCI system based on electrooculography (EOG) is proposed in this study. The proposed classification algorithm provides eye-state detection, including the fixation, saccade, and blinking states. Moreover, this algorithm can distinguish among ten kinds of saccade movements (i.e., up, down, left, right, farther left, farther right, up-left, down-left, up-right, and down-right). In addition, we developed an HCI system based on an eye-movement classification algorithm. This system provides an eye-dialing interface that can be used to improve the lives of people with disabilities. The results illustrate the good performance of the proposed classification algorithm. Moreover, the EOG-based system, which can detect ten different eye-movement features, can be utilized in real-life applications.

Subgroup label ranking aims to rank groups of labels using a single ranking model, is a new problem faced in preference learning. This paper introduces the Subgroup Preference Neural Network (SGPNN) that combines multiple networks have different activation function, learning rate, and output layer into one artificial neural network (ANN) to discover the hidden relation between the subgroups’ multi-labels. The SGPNN is a feedforward (FF), partially connected network that has a single middle layer and uses stairstep (SS) multi-valued activation function to enhance the prediction’s probability and accelerate the ranking convergence. The novel structure of the proposed SGPNN consists of a multi-activation function neuron (MAFN) in the middle layer to rank each subgroup independently. The SGPNN uses gradient ascent to maximize the Spearman ranking correlation between the groups of labels. Each label is represented by an output neuron that has a single SS function. The proposed SGPNN using conjoint dataset outperforms the other label ranking methods which uses each dataset individually. The proposed SGPNN achieves an average accuracy of 91.4% using the conjoint dataset compared to supervised clustering, decision tree, multilayer perceptron label ranking and label ranking forests that achieve an average accuracy of 60%, 84.8%, 69.2% and 73%, respectively, using the individual dataset.

There are current limitations in the recording technologies for measuring EEG activity in clinical and experimental applications. Acquisition systems involving wet electrodes are time-consuming and uncomfortable for the user. Furthermore, dehydration of the gel affects the quality of the acquired data and reliability of long-term monitoring. As a result, dry electrodes may be used to facilitate the transition from neuroscience research or clinical practice to real-life applications. EEG signals can be easily obtained using dry electrodes on the forehead, which provides extensive information concerning various cognitive dysfunctions and disorders. This paper presents the usefulness of the forehead EEG with advanced sensing technology and signal processing algorithms to support people with healthcare needs, such as monitoring sleep, predicting headaches, and treating depression. The proposed system for evaluating sleep quality is capable of identifying five sleep stages to track nightly sleep patterns. Additionally, people with episodic migraines can be notified of an imminent migraine headache hours in advance through monitoring forehead EEG dynamics. The depression treatment screening system can predict the efficacy of rapid antidepressant agents. It is evident that frontal EEG activity is critically involved in sleep management, headache prevention, and depression treatment. The use of dry electrodes on the forehead allows for easy and rapid monitoring on an everyday basis. The advances in EEG recording and analysis ensure a promising future in support of personal healthcare solutions.

Exposure to air pollutants is known to have adverse effects on human health; however, little is known about the association between hydrocarbons in air and an ischemic stroke (IS) event. We investigated whether long-term exposure to airborne hydrocarbons, including volatile organic compounds, increased IS risk. This retrospective cohort study included 283,666 people aged 40 years or older in Taiwan. Cox proportional hazards regression analysis was used to fit single- and multiple-pollutant models for two targeted pollutants, total hydrocarbons (THC) and nonmethane hydrocarbons (NMHC), and estimated the risk of IS. Before controlling for multiple pollutants, hazard ratios (HRs) of IS with 95% confidence intervals for the overall population were 2.69 (2.64-2.74) at 0.16-ppm increase in THC and 1.62 (1.59-1.66) at 0.11-ppm increase in NMHC. For the multiple-pollutant models controlling for PM2.5, the adjusted HR was 3.64 (3.56-3.72) for THC and 2.21 (2.16-2.26) for NMHC. Our findings suggest that long-term exposure to THC and NMHC may be a risk factor for IS development.

Recently, there has been an increase in the production of devices to monitor mental health and stress as means for expediting detection, and subsequent management of these conditions. The objective of this review is to identify and critically appraise the most recent smart devices and wearable technologies used to identify depression, anxiety, and stress, and the physiological process(es) linked to their detection. The MEDLINE, CINAHL, Cochrane Central, and PsycINFO databases were used to identify studies which utilised smart devices and wearable technologies to detect or monitor anxiety, depression, or stress. The included articles that assessed stress and anxiety unanimously used heart rate variability (HRV) parameters for detection of anxiety and stress, with the latter better detected by HRV and electroencephalogram (EGG) together. Electrodermal activity was used in recent studies, with high accuracy for stress detection; however, with questionable reliability. Depression was found to be largely detected using specific EEG signatures; however, devices detecting depression using EEG are not currently available on the market. This systematic review highlights that average heart rate used by many commercially available smart devices is not as accurate in the detection of stress and anxiety compared with heart rate variability, electrodermal activity, and possibly respiratory rate.

Internet of Things is smartly changing various existing research areas into new themes, including smart health, smart home, smart industry, and smart transport. Relying on the basis of “smart transport,” Internet of Vehicles (IoV) is evolving as a new theme of research and development from vehicular ad hoc networks (VANETs). This paper presents a comprehensive framework of IoV with emphasis on layered architecture, protocol stack, network model, challenges, and future aspects. Specifically, following the background on the evolution of VANETs and motivation on IoV an overview of IoV is presented as the heterogeneous vehicular networks. The IoV includes five types of vehicular communications, namely, vehicle-to-vehicle, vehicle-to-roadside, vehicle-to-infrastructure of cellular networks, vehicle-to-personal devices, and vehicle-to-sensors. A five layered architecture of IoV is proposed considering functionalities and representations of each layer. A protocol stack for the layered architecture is structured considering management, operational, and security planes. A network model of IoV is proposed based on the three network elements, including cloud, connection, and client. The benefits of the design and development of IoV are highlighted by performing a qualitative comparison between IoV and VANETs. Finally, the challenges ahead for realizing IoV are discussed and future aspects of IoV are envisioned.

This paper proposes a novel approach for selecting a subset of features in semi-supervised datasets where only some of the patterns are labeled. The whole process is completed in two phases. In the first phase, i.e., Phase-I, the whole dataset is divided into two parts: The first part, which contains labeled patterns, and the second part, which contains unlabeled patterns. In the first part, a small number of features are identified using well-known maximum relevance (from first part) and minimum redundancy (whole dataset) based feature selection approaches using the correlation coefficient. The subset of features from the identified set of features, which produces a high classification accuracy using any supervised classifier from labeled patterns, is selected for later processing. In the second phase, i.e., Phase-II, the patterns belonging to the first and second part are clustered separately into the available number of classes of the dataset. In the clusters of the first part, take the majority of patterns belonging to a cluster as the class for that cluster, which is given already. Form the pairs of cluster centroids made in the first and second part. The centroid of the second part nearest to a centroid of the first part will be paired. As the class of the first centroid is known, the same class can be assigned to the centroid of the cluster of the second part, which is unknown. The actual class of the patterns if known for the second part of the dataset can be used to test the classification accuracy of patterns in the second part. The proposed two-phase approach performs well in terms of classification accuracy and number of features selected on the given benchmarked datasets.