To address the challenge of dynamically adjusting HUD illuminance in complex driving environments where ambient light, weather, and driving tasks collectively affect HUD readability, this study designed a driving simulator-based controlled experiment. The experiment investigated variations in HUD optimal perceived luminance and eye activity patterns across six operational scenarios. These scenarios were constructed by combining three weather conditions (clear, rainy, foggy) with two driving tasks (straight-line navigation, lane-changing) and were tested under eight ambient illumination levels that cover typical real-world driving environments. Twenty participants were recruited to participate in the experiment. Results revealed that the optimal perceived illuminance of HUDs differed extremely significantly across the three weather conditions: foggy weather required the highest HUD brightness, followed by rainy weather, while clear weather required the lowest. Based on the finding and the experimental parameter settings, a logarithmic model describing the relationship between the optimal perceived illuminance of HUDs and ambient illumination was established. This study's contribution lies in providing references for the interactive design of HUD brightness adjustment under different weather conditions, thereby contributing to the enhancement of driving safety.

Bare-hand interaction techniques enable users to directly manipulate virtual objects through virtual hands in virtual reality systems, exhibiting distinct perceptual-motor coupling characteristics. However, the absence of tactile feedback and the inherent visual uncertainty increase the complexity of hand-movement control in such environments. This study examined how two virtual button interface conditions with different attentional-guidance characteristics, namely goal-based and stimulus-based interfaces, are associated with differences in hand-targeting control performance in virtual button selection tasks. A series of kinematic metrics were evaluated using generalized linear mixed models to assess statistical differences between conditions. Additionally, overshoot and undershoot tendencies as well as ballistic-corrective phase segmentation were analyzed, and movement trajectories were visualized to characterize the underlying hand-movement control patterns. The results indicated that movements performed under the goal-based condition exhibited prolonged initiation latency, lower initial velocity peaks, a greater tendency toward undershooting, and more frequent corrective trajectory adjustments, suggesting greater online control demands during movement execution. In contrast, movements under the stimulus-based condition, supported by salient visual cues, were associated with shorter task completion times, faster movement profiles, more pronounced overshoot tendencies, and a more distinct ballistic phase, suggesting a more open-loop-like movement pattern. Furthermore, the study proposes strategies for optimizing virtual interface design and enhancing interaction intelligence. These insights provide practical implications for VR applications in industrial operations, medical training, skill acquisition, and collaborative work environments.

In complex driving scenarios with multi-source risks, effective attentional guidance is essential for drivers to regain situational awareness. This study investigates how attentional cueing strategies embedded in takeover requests (TORs), and varying environmental visibility levels influence driver decision-making during automation-to-human transitions. Forty-eight participants drove in a simulator under three visibility conditions (clear, light fog, dense fog) and received either action-oriented or hazard-oriented cues, plus a no-cue baseline. Dependent measures included takeover time, maximum deceleration, standard deviation of lateral offset, time to collision, eye movement metrics, and subjective assessments. Overall, the threat-cued strategy supports autonomous decision-making by enhancing situational awareness, whereas the action-cued strategy improves takeover quality by providing intention-aligned guidance under limited visibility, though it may entail a potential risk of automation complacency.

Automated retrieval in large-scale remote sensing datasets often becomes unstable for occluded or ambiguous targets. This study proposes a brain-computer collaborative retrieval system in which a Rotation-equivariant Detector first screens images and delegates only medium-confidence detections to human verification through a rapid serial visual presentation paradigm. Behavioral experiments identified standardized image size, gray background, edge cues, and 350 ms exposure as performance-optimized presentation settings. EEG analysis revealed robust P300 responses over posterior parietal-occipital electrodes, supporting single-trial discrimination of ambiguous targets. A physiological-data dual-driven decoding framework combined P300-guided trial cleanup, stochastic augmentation, and a Riemannian ensemble architecture to reduce label noise, temporal jitter, and inter-subject variability. The system achieved a mean AUC of 0.8003 for ambiguous target detection, indicating the feasibility of implicit neural feedback for reliable hard-case retrieval in remote sensing.

Reinforcement Learning from Human Feedback (RLHF) and related alignment pipelines hinge on high-quality human feedback data. However, many current "black-box" interfaces often trigger cognitive misalignment between humans and AI, leading to noisy feedback. Drawing on Shared Mental Model (SMM) theory, we propose transparency-based interfaces to align human-AI cognition by externalizing AI internal states. We conducted a 2 & times; 3 within-subjects experiment (N = 30) to examine how interface support (Baseline vs. SMM-supported) interacts with interaction granularity (Choice-only, Structured Annotation, Conversational Guidance). Behavioral and eye-tracking results revealed a "Cost of Transparency": although SMM support slightly increased early information engagement, it reduced decisional hesitation during task execution. Subjective measures further suggested that transparency reduced perceived cognitive burden rather than imposing a net cost, as reflected in lower mental-demand ratings. Furthermore, we identified structured annotation as the interaction "sweet spot," effectively balancing user agency with cognitive effort. These findings suggest that interfaces for human-in-theloop alignment tasks should function as cognitive scaffolds, externalizing AI uncertainty and decision-relevant cues to transform user behavior from hesitant guessing into informed decision-making. More broadly, the study shows that effective human feedback depends not only on model capability, but also on interface support for shared understanding.

Complex information system interfaces typically present high information density and diverse visual elements, making color configuration a critical factor influencing visual search efficiency and attentional allocation. While previous studies have examined individual visual variables in isolation, the combined effects of hue quantity and color contrast in complex interface contexts remain insufficiently understood. This study systematically investigates visual search in complex information interfaces using a repeated-measures experimental design, focusing on the effects of hue quantity and color contrast on user performance. Behavioral reaction time and eye-tracking measures were jointly analyzed to examine how interface color characteristics influence visual search across different processing stages, including overall task completion and initial attentional orienting. The results show that, at the behavioral level, color contrast exerts a stable facilitating effect on visual search efficiency, whereas the influence of hue quantity is relatively limited. However, during the initial orienting stage of visual search, the effect of contrast is strongly dependent on the level of hue quantity. Specifically, under low hue quantity conditions, increasing contrast significantly accelerates first entry into the target region; under medium hue quantity conditions, the facilitating effect of contrast diminishes; and under high hue quantity conditions, high contrast no longer facilitates—and may even delay—initial attentional orienting. These findings indicate that contrast is not a universally effective saliency design strategy, but one whose effectiveness is constrained by overall color complexity. By revealing stage-specific and condition-dependent interaction mechanisms between hue quantity and contrast, this study provides empirically grounded guidance for color configuration design in complex information system interfaces.

The integration of Large Language Models (LLMs) into product design has been reshaping professional workflows, yet the factors driving designers' adoption remain underexplored. This study identifies the key determinants of product designers' intention and actual use of LLMs by employing a mixed-method approach that combines the UTAUT2 model with Structural Equation Modeling (SEM) and Artificial Neural Network (ANN) analysis of data from 234 product designers and students. The SEM results revealed that Performance Expectancy (PE), Social Influence (SI), and Effort Expectancy (EE) influence Behavioral Intention (BI), while Facilitating Conditions (FC) and Habit (HB) directly impacted actual Use Behavior (UB). The ANN analysis provided a nuanced importance ranking for intention (SI > PE > EE) and for actual use (HB > FC), revealing that adoption was a process of passive adaptation driven by social pressure, while sustained use is primarily habit-driven. A counter-intuitive finding was the negative impact of EE on BI, attributed to the tension between a desire for easy tools and the fear of eroding professional identity. The findings offered actionable implications for developing and promoting LLMs that align with the Collaboration, Performance Pursuit and Professional Identity nature of Product Design.

With the increasing reliance on data-driven decision-making, information visualizations have become critical tools across various domains. However, cognitive biases frequently distort users’ interpretation of visualized data, leading to suboptimal outcomes. While prior research has documented cognitive biases in specific visualization contexts, most studies remain at the behavioral level and lack a unified framework for systematic bias modification. This study proposes a novel framework for understanding and modifying cognitive biases in information visualizations using event-related potentials (ERP). The framework consists of four key stages: 1) identification of potential cognitive bias triggers, 2) neural and behavioral characterization of cognitive biases, 3) preliminary design plans for cognitive bias modification, and 4) design evaluation and design strategy optimization. A case study on automotive dashboard interfaces demonstrates the framework’s effectiveness in addressing anchoring bias. Experimental results show that combined angular and textual encoding induces anchoring bias, as evidenced by specific ERP components (P2, P300, N400). Three design strategies were derived: context-appropriate textual warnings, vertical alignment of elements, and semantically consistent information positioning. This work bridges theoretical insights with practical applications, offering a structured and actionable guidance for creating bias-modified visualizations.

In interface design for high-risk environments, icon visual search performance is critical for system reliability and operational safety. Existing research has mainly focused on individual icon features, neglecting the effects of icon composition style and hue consistency from an icon set perspective, as well as the interference effects of specific noise levels in real-world environments. This study integrated behavioral experiments and eye-tracking technology to investigate how composition style (planes vs. lines), hue consistency (consistent vs. inconsistent), and noise level (45 dB vs. 60 dB vs. 75 dB) affect users' visual search performance from an icon set perspective. Supplementary exploratory analyses based on heat maps, track maps, and semi-structured interviews were conducted to aid interpretation of the quantitative findings. A 2 & times; 2 & times; 3 within-subjects repeated-measures design was employed, and 30 participants completed a total of 17,280 icon search trials. Behavioral and eyetracking results showed that neither the interaction effect between composition style and hue consistency nor the main effect of noise level reached statistical significance. However, significant interaction effects between composition style and noise level were observed for both task completion time (TCT) and fixation count (FC). A comprehensive evaluation indicated that, at 60 dB, the icon set combining plane-based icons with consistent hue showed more favorable visual search performance than the other three icon set types. In contrast, the icon set combining line-based icons with inconsistent hue showed the least favorable visual search performance across the three noise levels. These findings offer preliminary implications for icon set design in high-risk digital interfaces and for future research on visual search under environmental noise.

Viewport transitions are common in complex information systems, yet their associated cognitive costs remain underexplored. This study investigates how visual information density modulate attentional reorientation demands during viewport transitions. Using a dual-task paradigm involving visual enumeration and auditory discrimination, temporal patterns of auditory perceptual sensitivity, response bias and auditory response time were analysed as indirect indicators of residual attentional capacity under visual load. Results indicated that current viewport density played a central role in shaping viewport switching costs. Transitions to low-density viewports were associated with more efficient attentional reorientation, whereas transitions to high-density viewports imposed greater early reorientation demands. In contrast, inter-viewport density difference did not show statistically robust effect, although descriptive patterns suggested a possible advantage for stable low-density transitions. These findings provide an initial, pattern-based account of how information density influences viewport switching cost and offer preliminary design implications for visual information density management in multi-viewport interfaces.

Artificial intelligence (AI) aids many open innovation platforms in human ideation and evaluation. While current research highlights AI's innovation potential, it has paid less attention to the ethical challenges from information flow in human-AI collaboration. This flow causes a tradeoff between generating novel ideas and ensuring their accurate evaluation free from strategic manipulation. Using an evolutionary game model for open innovation's design and evaluation stages, this study explores HAI reciprocity's impact on interpersonal dynamics and the interplay between platform incentives and AI. Results show constraints are crucial to sustaining positive cycles of HAI reciprocity. Unconditionally cooperative AI slightly boosts group innovation but reduces individual human performance and hinders valid evaluation. In contrast, Reputation-based AI improves both human and population performance. With fixed incentives, having more AI increases system stability and improves innovation outcomes. Platforms should integrate AI into mechanism design, rather than treating it as a technical tool.

This study applies the Kano model theory to community aging-friendly service requirements and conducts an in-depth excavation of the community service requirements of the elderly in urban communities to form a multi-level demand indicator system. In recent years, community endowment combines the advantages of family care and institutional care and has gained general recognition from the elderly for its convenience and comfort. However, many administrators focus on construction only and neglect the requirements of the elderly and user experience, resulting in vast wasted or insufficient service resources, so it is important to clarify the requirements of the elderly on community aging-friendly services. However, the community endowment model emerged relatively late in China, and research on the community service requirements of senior citizens is still immature. Therefore, to promote the development of community endowment, the requirements of the elderly should be considered first. In this study, we investigated the daily activities and community service requirements of the elderly, disassembled the relevant resources covered by specific community services, and summarized the quality attributes of the service, so as to design the Kano questionnaire and construct the Kano model of the quality attributes of aging-friendly service resources.

Based on the SEEV model, the nuclear power plant monitoring and display interface is optimized and designed. First, we summarize the key contents and existing problems of the interface, then introduce the salience and effort factors in the SEEV model to optimize the design, and finally use eye tracking methods to study the old and new interfaces. Combined with the analysis of eye movement experimental data, the visual cognitive characteristics of subjects performing visual search tasks in different interfaces were obtained. Experimental results show that combining the SEEV model can better guide subjects' attention allocation strategies, thereby improving work efficiency.

With the increasingly widespread application of flying robots in various spatial environments, their auxiliary role in collaborative tasks has become more prominent. These tasks often necessitate close interaction with humans. Consequently, designers must thoroughly consider how humans perceive and interpret the behavior of flying robots. However, existing research has yet to accurately represent all motion states of flying robots. This study aims to explore the design space for enabling flying robots to communicate their flight intentions to nearby users. It proposes a set of lighting language rules, utilizing eight lights, to effectively convey the intentions of flying robots and precisely represent their motion states. Firstly, this study systematically decomposed and analyzed the motion states of flying robots in operational scenarios, identifying a total of 14 key motion states. Based on this analysis, design interviews were conducted to summarize design rules, and three schemes of lighting signal schemes were developed for the flying robots. Subsequently, through user studies, the effectiveness of each scheme of signal schemes was evaluated, and participants were required to predict the intentions of the robots. The results demonstrated that the proposed design significantly improved observers’ response speed and accuracy by 10.7%, while reducing their mental workload by 31.1%. The design scheme significantly enhanced users' ability to understand and predict the flight intentions of robots. The summarized design principles and design schemes laid an important foundation for the promotion and application of future flying robots in various operational contexts.

With the acceleration of the aging process and the increase of the elderly population, the construction of age-friendly communities has become an important link in the development of elderly care services. Literature collection and collation are conducted to analyze the progress of community service facility system construction on age-friendly community and community service facility system construction, to elaborate on the relevant research results. It summaries the current status and hot issues of age-friendly communities, analyzes and summarizes the related research on the construction of the age-friendly community service facility system from four dimensions: the configuration of community facilities for age-friendly, the spatial layout of facilities for age-friendly, the transformation ideas of facilities for age-friendly, and the evaluation of age-friendly community. The overall ergo-nomic design of community service provides an interactive community service map design scheme that can adapt to the characteristics and needs of the daily behavior and activities of the elderly. In the future, more attention should be paid to the high-level needs of the elderly and the construction of guiding principles and evaluation indicators for the age-community.

Conceptual design is inherently a social and creative activity. Most studies on conceptual design of designer teams focused primarily on behavioral aspects, leaving cross-brain coupling neural mechanisms underlying designer teams' collaboration unexplored. Investigating inter-brain synchrony (IBS) offers a critical perspective on how shared neural activity supports key collaborative processes, such as coordination, communication, and team creativity. This study investigated the effects of design interaction modes (face-to-face vs. remote virtual vs. electronic brainstorming; FTF vs. RV vs. EBS) on designer teams' interactive behaviors and IBS during conceptual design. Using fNIRS-based hyperscanning, neural activities in the right prefrontal cortex and right temporoparietal junction (r-TPJ) were recorded for 72 designers (36 dyads), and behavioral characteristics, IBS, and temporal dynamics of these metrics across modes were analyzed. Results showed that FTF teams outperformed RV and EBS in creative design performance, cooperation level, team flexibility, perspective-taking, and turn- taking. Creative design performance and cooperation level increased over time across all modes, particularly in FTF and RV, while team flexibility, perspective-taking, and turn-taking initially rose before declining, notably in FTF and RV. fNIRS data revealed greater IBS in r-TPJ and between r-TPJ and right dorsolateral prefrontal cortex (r-DLPFC) in RV compared to FTF and EBS, both following a U-shaped temporal trend. Cooperation level, perspective-taking, and turn-taking positively correlated with RIBS in r-TPJ, while cooperation level correlated with RIBS between r-TPJ and r-DLPFC. These results highlight distinct behavioral and neural synchronization patterns across interaction modes in designer teams during conceptual design process, with FTF mode performing best. These findings enhanced understanding of designer teams' interactive cognition, contributed to design neurocognition research, and offered practical implications for designing tools and training programs to optimize team performance.

This study explores the impact of waterfront interface lighting on visual behavior using eye-tracking data and virtual reality, focusing on the Xuanwu Lake waterfront in Nanjing, China. We examine how various brightness combinations of focal building, core buildings, secondary buildings, city walls, and vegetation affect fixation counts and durations. Thirty-one participants viewed 130 panoramic images, and eye-tracking data were collected using Varjo XR-3 mixed reality devices. The results show that the brightness of focal and core buildings significantly affects fixation counts, while the city wall's brightness and its interaction with other lighting parameters influence fixation duration. Secondary buildings had little impact due to their lower brightness. Lighting configurations with higher contrast and multi-layered brightness schemes increased fixation time, suggesting that well-designed lighting can attract more attention and enhance visual experience. This study provides data-driven insights for optimizing urban waterfront lighting design, helping create more engaging and visually appealing night-time environments to support the nighttime economy.

This paper explores the perceptual quality of visual coding designs for 3D charts in data visualization. With the increasing use of 3D-style interfaces, 3D charts such as pie charts and bar charts are gaining popularity due to their enhanced visual appeal and higher user experience. However, the design guidelines for 3D chart visual coding, particularly regarding transparency, viewing angles, and category quantities, remain under explored. Inappropriate design choices in these areas can lead to increased cognitive load, complicating data interpretation. This study aims to assess optimal visual coding strategies for the typical 3D pie chart through a behavioral experiment, offering recommendations for design to reduce cognitive strain and improve overall visual quality.

This study investigates the role imagery preferences of medical virtual agents (VAs) for elderly users, presenting a prompt engineering (PE) framework to enhance emotional resonance and user acceptance. Through a mixed-methods approach, we developed the S-MVA (Medical Virtual Agents for Seniors) prompts model, generated six AI-driven avatars, and evaluated their emotional impact on 100 elderly participants in China by using the PrEmo2 scale. Results reveal significant differences in emotional responses across avatar designs, with avatars featuring anthropomorphic warmth (e.g., gentle smiles, family-like traits) and culturally tailored elements (e.g., lab coats, polite communication) eliciting stronger positive emotions. This research reveals the emotional effectiveness of tailored generative design in fostering trust and engagement among older adults, providing a structured framework for generating medical VAs imagery that address their emotional needs.