In today’s academic scenario, understanding learners’ emotional responses during academic learning is important to improve learning ability. This study provides a comparative analysis of facial emotions from interactions with both real human (RH) and virtual human (VH) in the context of online academic learning. Facial video data were collected from participants engaged in both RH and VH learning sessions. Facial landmarks were extracted using the MediaPipe Face Mesh model and six emotional states were mapped from computed action unit (AU) scores. A convolutional neural network (CNN) was trained on the FER-2013 and extended CK+ datasets to classify six facial emotional states from the acquired dataset. Emotion intensity was computed based on AU scores for each detected state. Results revealed that happiness and surprise intensities were significantly higher during VH interactions compared to RH. An ANOVA test confirmed statistically significant differences in emotional intensity between RH and VH interactions.

Virtual Reality (VR) provides immersive and interactive experiences in healthcare, education, entertainment, and defense. However, cybersickness remains a major barrier to its widespread adoption, reducing user comfort and engagement. Early and accurate detection of cybersickness is critical to developing adaptive VR systems that ensure safety and improve usability. In this study, we propose a novel real-time cybersickness detection approach using Bidirectional Long Short-Term Memory (Bi-LSTM) networks trained on electroencephalography (EEG) signals. Power Spectral Density (PSD) and Signal Magnitude Area (SMA) features were extracted to capture frequency- and amplitude-related characteristics of cybersickness. EEG data were collected from six electrodes across frontal (F3–F4), prefrontal (FP1–FP2), and central parietal (P3–P4) regions during VR exposure. The proposed Bi-LSTM model achieved 95% classification accuracy, significantly outperforming baseline methods. Results indicate that cybersickness can be reliably detected with a compact EEG setup, supporting resource-efficient, real-time monitoring for adaptive VR environments.

Personal intelligent agents (IAs) are increasingly embedded in everyday life, a trend accelerated by generative AI technologies. Despite their growing presence, these agents often remain fragmented across different life domains and environments. This workshop explores how to design integrated IA ecosystems emphasizing continuity, coordination, and human-centered values. Participants with varied perspectives will collaboratively develop frameworks, scenarios, and guidelines for cohesive personal agent systems that enrich user experiences holistically. By examining factors that shape users' preferences for information integration or separation, we aim to inform the design of coherent, user-aligned multi-agent systems.

In-vehicle intelligent agents (IVIAs) have been developed to improve user experience in autonomous vehicles. Yet, the impact of the automation system on driver behavior and perception toward IVIAs is unclear. In this study, we conducted three experiments with 73 participants in a driving simulator to examine how automation system parameters (the level of automation system and IVIA features) influence driver attitudes, cognition, and behaviors when driving or riding in a simulated vehicle. We focused on subjective evaluations of driver-agent interaction and driver trust toward IVIAs to assess driver attitudes, driver situation awareness, and visual distraction to capture their cognition, and their driving performance to understand their behaviors. Our results show that the level of automation system affects drivers’ attitudes toward agent capabilities (e.g. perceived intelligence). Embodiment benefits are more pronounced with Level 5 systems, while speech style, in general, is more influential in determining affective aspects of user attitudes (e.g. Warmth, Likability). As the level of automation increases, drivers engage in more visual distractions. In addition, conversational speech style in general encouraged safer driving behaviors indicated by more stable lateral control under lower levels of automation. Our findings uncover the path of how system parameters affect driver behaviors through system evaluation and trust in agents. These findings have important implications for the development of cohesive user experiences in future transportation systems.

The diffusion of virtual reality (VR) technology has highlighted several user experiences challenges, including cybersickness (CS), mental workload (MWL), eye fatigue (EF), and physical fatigue (PHF). To address these issues, a comprehensive human factors and ergonomics (HFE) tool is necessary. This study developed a specialized HFE questionnaire to assess primary issues in VR environments. Ninety-three participants used VR headsets to test the questionnaire, which was validated through expert opinions and statistical analysis. The questionnaire consists of 19 questions categorized into four groups: MWL (4 items), PHF (5 items), CS (5 items), and EF (5 items). The questionnaire showed high reliability and validity, with a Cronbach’s alpha coefficient of 0.91, a mean content validity index of 0.83, and a content validity ratio of 0.81. Structural validity was also confirmed with acceptable Chi-square and root mean square error of approximation scores. The results support the validity and reliability of the questionnaire, making it a useful tool for assessing and improving VR user experiences.

This study examined the effects of two locomotion methods (joystick and teleportation) and two postures (sitting and standing) on mental workload (MWL) in virtual reality (VR). Sixty participants played a VR game using a 2 × 2 experimental design, with assessments including Electroencephalography (alpha and theta bands), Electrocardiography (heart rate and heart-rate variability indices), and the NASA Task Load Index (NASA-TLX). The results showed lower theta activities and higher alpha activities in the joystick condition, indicating higher MWL in the teleportation condition, primarily due to time demand and effort. Sitting posture resulted in lower mental load and higher heart rate with increased Standard Deviation of Normal-to-Normal intervals (SDNN) compared to standing, suggesting higher MWL in the standing posture. The NASA-TLX highlighted physical demand, effort, and time demand as crucial factors in MWL related to posture. These findings provide a basis for developing human factors and ergonomics (HF/E) guidelines for VR, emphasizing the importance of locomotion methods and user posture in reducing mental workload.

This study investigated the main causes of Air Force pilot fatigue, which could lead to aviation accidents. Data were collected from Korean pilots through in-person survey, utilizing the Human Factors Analysis and Classification System (HFACS). The study focused on preconditions for unsafe acts and unsafe supervision, as well as detecting fatigue using a modified checklist. A three-factor structure comprising 17 subfactors was identified, exhibiting robust psychometric properties, good model fit and reliability. Strong positive relationships were observed between pilot fatigue and all subfactors. The study highlighted the influence of age, job identity, and total flight hours on pilot fatigue. Instructors and more experienced pilots had higher fatigue scores than younger pilots. These findings underscore the need for targeted interventions to reduce fatigue among middle-aged pilots, instructors, and experienced pilots. The results validate the model’s applicability in Korea and suggest its potential utility in other industrialized nations.

Artificial intelligence-powered coding assistants (AI-CAs) have become essential tools in programming; however, there is limited understanding of the mechanisms driving programmers’ adoption of these tools in their daily coding tasks. This study aims to examine the role of utilitarian and hedonic values in the adoption of AI-CAs by extending the Technology Acceptance Model (TAM). The data gathered from an online survey of 283 Korean programmers is analyzed using structural equation modeling. The results showed that both perceived enjoyment and perceived usefulness positively influence the attitudes and usage intentions toward AI-CAs. Interestingly, perceived enjoyment has a stronger influence on the intention to use than perceived usefulness, suggesting that recognizing the intrinsic motivation for using AI-CAs is crucial for fully leveraging their benefits. The model also confirms that the compatibility and relative advantages of AI-CAs enhance their adoption. This research enriches the current knowledge base by incorporating hedonic values into the TAM, offering new insights on the design of AI-CAs aimed at enhancing their adoption among developers.

A variety of human factors/ergonomics (HFE) problems have been studied by researchers and developers in VR environments. This systematic review aimed to summarize important HFE issues and classify the validated instruments used to quantify them in virtual reality environments. The most representative electronic databases for this review (2013–2022) were searched for original articles. The results showed that aspects, such as cybersickness, visual fatigue, mental workload, performance, spatial presence, and usability were the most relevant HFE issues assessed, whereas some aspects, such as physical workload, posture, stress, and discomfort, were consider less often. Previous studies have neglected some human factors and ergonomic issues, such as physical ergonomics, stress, and aftereffects, such as fatigue and human error. In virtual environments, presence was an emerging human factor compared to real environments. Most techniques were unidimensional and subjective. Future studies should focus on more factors and risks associated with HFE by emphasizing objective techniques and multidimensional subjective methods.

This study investigated the effects of the target size and movement distance on user performance and workload in a virtual reality (VR) environment. In a repeated-measures laboratory study, 36 participants (18 male and 18 female) performed the drag-and-drop task as a standard human–computer interaction (HCI) task with different target sizes (1, 1.5, 2, 2.5, and 3 cm) and movement distances (5, 9, 13, 17, and 20 cm). Task completion time (TCT), error rate, and movement time (MT) were measured as performance indices, whereas physical load and effort were assessed as workload indices. The results demonstrated that the target size and movement distance significantly affected all performance measures and workload indices. Large target sizes produced better performance and lower workloads; however, large movement distances decreased performance and increased workload. However, sex had no significant effect on the performance or workload during the drag-and-drop tasks. The best target sizes were 2.5 and 3 cm, and the worst size was 1 cm. The best movement distances were 5 and 9 cm, and the worst distance was 20 cm. The results of this study can provide useful reference information for developing VR technology based on human factors and demonstrate that additional basic research is required to reflect the distinctive features of VR in the future.

Virtual reality (VR) users experience unwanted symptoms, such as body imbalance, nausea, dizziness, and loss of presence. This study aims to investigate the effects of posture and locomotion on postural stability, cybersickness, and presence in VR environments. Twenty participants played a VR game under different conditions depending on posture (standing and sitting) and locomotion methods (joystick and teleportation), and seven dependent variables (COM, AP displacement, ML displacement, ASL, PSQ, VRSQ, and SOP) were analyzed to observe postural stability, cybersickness, and presence. The results revealed that postural instability increased when the task was performed in the standing posture with the joystick locomotion method compared to other conditions. The AP displacement, ML displacement, and ASL were lower under teleportation conditions. The PSQ scores indicated that postural stability was better in the sitting posture than in the standing posture. The VRSQ score revealed that the sitting with teleportation condition had less cybersickness than the other conditions. The SOP score was the highest in the standing posture with the teleportation condition. This study concludes that a sitting posture with teleportation locomotion can be considered when designing games in which users actively interact with virtual movements.

With a growing interest in head-mounted display (HMD)-based virtual reality (VR) environments, there have been lots of studies enhancing user experience (UX) in the context. In particular, the study of motion sickness (MS) symptoms, which are a major barrier to providing a positive UX, is of high importance. This study investigated the effects of degree of freedom (DOF) and field of view (FOV) of HMD on MS symptoms. A user experiment was designed based on a 2 × 2 mixed design with DOF as a between-subject design variable (3-DOF and 6-DOF) and FOV as a within-subject design variable (Narrow and Wide). Participants experienced VR game content in four conditions in random order. MS symptoms were captured using heart rate variability (HRV) and virtual reality sickness questionnaire (VRSQ) at the pre- and post-task moment. The results showed that MS symptoms occurred more in the 3-DOF condition than in the 6-DOF. Further, MS symptoms in DOF conditions were adjusted depending on the FOV. In the Wide condition, we found a significant difference in MS symptoms depending on the DOF, but in the Narrow condition, no significant differences were found. Through this finding, we were able to not only show the effectiveness of HRV as MS measures but also provide meaningful design insights into HMD-based VR practices.

Industry 4.0 is potentially innovative in the workers’ role, which is becoming increasingly involved in smart activities. In this situation, it is necessary to improve highly repetitive uncomfortable working postures to reduce physical risks. This study intends to assess physical risks during VR interaction for manufacturing work. Posture-related physical risk levels were calculated using ergonomic risk assessment tools RULA, REBA, and OWAS. Three task conditions were considered for the experiment in a VR-based car-assembly environment. An analysis of variance was applied to investigate significant differences between task conditions, and it suggested that a higher risk level was obtained while working in the overhead position for RULA and REBA, whereas the squatting position obtained a higher risk level for OWAS. Sensitivity analysis identified that the upper arm and neck were responsible for the highest risk level for RULA, the upper arm, neck, and trunk for REBA, and the back posture parameter for OWAS.

Accepted version

The introduction of fully automated vehicles (FAVs) will change user experiences (UX) in personal transportation. In order for FAVs to become a life enhancing technology, it is required to design vehicular applications and user interfaces based on users’ expectations. To this end, we investigated user needs and design requirements. First, we elicited design taxonomy and use cases through literature review and trend analysis. Using these materials, expert interviews (N = 9) and focus group interviews (N = 10) were conducted. Through the qualitative analysis, we obtained twelve categories of user needs and devised design requirements based on the updated design taxonomy. While some of them have been an extension of current experiences in manual driving, completely new demands have also emerged within FAVs. Our findings contribute to designing UX in FAVs by satisfying users’ expectations and key values that can guide designers.

Speech style and embodiment are two widely researched characteristics of in-vehicle intelligent agents (IVIAs). This study aimed to investigate the influence of speech style (informative vs. conversational) and embodiment (voice-only vs. robot) and their interaction effects on driver-agent interaction. We conducted a driving simulator experiment, where 24 young drivers experienced four different fully autonomous driving scenarios, accompanied by four types of agents each, and completed subjective questionnaires about their perception towards the agents. Results showed that both conversational agents and robot agents promoted drivers' likability and perceived warmth. These two features also demonstrated independent impacts. Conversational agents received higher anthropomorphism and animacy scores, while robot agents received higher competence and lower perceived workload scores. The pupillometry indicated that drivers were more engaged while accompanied by conversational agents. Our findings are able to provide insights on applying different features to IVIAs to fulfill various user needs in highly intelligent autonomous vehicles.

This study aimed to investigate the effects of non-driving-related tasks (NDRTs) on takeover quality in the context of automated driving. Specifically, we examined the effects of three categories of NDRT attributes (i.e., physical, cognitive, and visual) on longitudinal and lateral driving measures when the drivers resumed control. We designed a driving simulator study where the participants experienced automated driving journeys and takeover situations. When the automated mode was activated, drivers engaged in one of the nine NDRTs. The results showed that the cognitive load of NDRTs had a significant negative correlation with both longitudinal and lateral control measures. However, the effects of two attributes in the physical category and one attribute in the visual category on driving performance did not show statistical significance. Overall, the findings indicated that the influence of cognitive attributes on takeover quality is more salient than that of the physical and visual attributes, which provides insights into the understanding of takeover situations to improve driving safety.

The objective of this study was to investigate the touch area that can be comfortably reached by the thumb during one-handed smartphone interaction. To achieve the research objective, we introduced the concept of natural thumb position when designing a tapping task and conducted an user experiment. The independent variables were the target distance and direction from the natural thumb position, and the three dependent variables were the task performance, information throughput, and touch accuracy. The results showed that participants performed the task comfortably in the diagonal direction between the upper right and the lower left side of the screen. The task performance deteriorated as the target distance increased, especially at 45 mm or more. The touch accuracy was measured using X- and Y-coordinates data. Participants touched the left side of the target center, except near the proximal area of the hand. They also touched the points above the center of the target in the upper screen area and points below the center of the target in the lower screen area. The findings of this study provided insights for designing a smartphone touch interface considering the comfortable touch areas of one-handed interaction.

Smart functions in vehicles have led to an increase in the complexity of control interfaces. This study aims to develop a model for evaluating in-vehicle controller complexity and to investigate the relationship between complexity and task performance. A research framework consisting of three complexity dimensions (functional, behavioral, and structural dimensions) and controller-related variables was developed based on previous literature. A user experiment was conducted using 10 vehicles and 91 participants. A regression analysis was used to examine the relationship between the measurement variables and perceived controller complexity, and the results indicated correlations between them. An increase in functional dimension variables caused an increase in the perceived complexity level, while behavioral dimension variables are not a statistically significant predictor. Structural dimension variables showed different results depending on the characteristics of the variables. The results of the control task experiment showed a negative correlation between task performance and the perceived complexity level. In addition, satisfaction decreased with increasing levels of complexity. These results provide insights for managing in-vehicle controller complexity.

Despite enhancements in the visual complexity of in-vehicle information display in recent years, few studies have examined the effects of such increased complexity. We conducted this study with the following objectives: (1) to suggest a framework for predicting the perceived visual complexity (PVC) of in-vehicle information display; (2) to examine the effects of PVC on the visual behavior of human operators; (3) to investigate the relationship between preferences and PVC. A theoretical framework to evaluate PVC was developed, and a survey study was used to collect participants’ perceptions on visual complexity. A regression analysis was employed to find the relationship between each of three factors and PVC. Two of the factors—quantity and variety—showed a positive correlation with PVC, whereas the third factor, relation, exhibited a negative correlation. Visual search experiments were conducted to test the effects of PVC on the performance of visual search tasks and glance behavior. The results showed that the high level of PVC leads to more time-on-task and number of fixations. We also found that preference for in-vehicle information displays was inversely proportional to PVC. The results enable us to predict how human operators perceive visual complexity and explain the influence of PVC on human behavior.