New Perspectives on Eye-Tracking: Theory, Methods, and Applications

1. Introduction

Eye-tracking technology has evolved into a cornerstone of modern behavioral and cognitive research, providing precise insight into how individuals perceive, interpret, and interact with their visual environments. Rapid progress in computer vision, artificial intelligence, and sensor miniaturization has transformed traditional gaze-tracking systems into robust analytical platforms capable of real-time monitoring and multimodal integration [1,2,3,4]. Recent transformer-based and semi-supervised architectures have markedly improved the accuracy of gaze-estimation pipelines [1,5,6], while deep learning has enabled joint modeling of eye landmarks, states, and visual context [2,6,7]. These developments expand the applicability of eye-tracking to a wide spectrum of domains, from clinical diagnostics [7,8,9,10,11] and neuroergonomics [12,13] to transportation safety [14,15,16,17,18] and immersive virtual environments [4,19,20,21,22].

In the biomedical sphere, gaze-pattern analysis and oculomotor metrics have become valuable tools for detecting visual and cognitive disorders. Studies have demonstrated the potential of fixation-based modeling for the early detection of dyslexia [8,9], schizophrenia [10], and diabetic retinopathy [23]. Deep-learning classifiers and recurrent neural networks increasingly interpret complex eye-movement data for cognitive screening [6,8,11,23]. Simultaneously, advances in neuro-optometric research have connected pupillometric dynamics with cortical plasticity and perceptual dominance [24,25], offering promising biomarkers for neurological adaptation and rehabilitation. Similar bio-signal fusion approaches, such as hybrid electrooculogram networks [7,11], reveal that eye-tracking can complement physiological sensing in healthcare monitoring.

In applied psychology and ergonomics, eye-tracking serves as a direct window into attention, situational awareness, and decision-making [12,13,14,15,16,17,18]. Studies focusing on pedestrians [14], drivers [15,16,17,18], and operators in complex human–machine interfaces reveal that gaze behavior encodes both task-load distribution and hazard anticipation. Combining gaze metrics with EEG or vehicle telemetry yields comprehensive measures of cognitive load and safety risk [12,13,15,17]. These insights have inspired new approaches to automation design, intelligent driver-assistance systems, and augmented interfaces that respond dynamically to user intent [19,20,21,22].

Parallel progress has emerged in virtual- and augmented-reality environments, where gaze data enrich interactivity and immersion [4,19,20,21,22,26]. Applications range from retail analytics and avatar-mediated communication [19,27] to clinical training and digital pathology [7,28]. The fusion of real-time gaze tracking with spatial computing enables adaptive rendering, attention-aware simulation, and context-sensitive learning systems [21,22,26]. As vision science, machine learning, and interface design converge, eye-tracking is transitioning from a diagnostic and observational tool to an active component of intelligent perception systems.

This Special Issue of Applied Sciences, entitled “Eye-Tracking Technologies: Theory, Methods and Applications,” was conceived to reflect this interdisciplinary momentum. It brings together contributions addressing both the methodological foundations and the practical implementations of contemporary eye-tracking research. The articles encompass computational modeling of gaze patterns, multimodal integration in human–machine systems, clinical assessment using oculomotor data, and application-driven innovations in immersive and assistive technologies. Collectively, these studies highlight the dual role of eye-tracking as a research instrument and as a core enabler of next-generation adaptive interfaces.

This Special Issue gathers contributions published between 2022 and 2024 from research groups across Europe, Asia, and the Americas. The papers span core methodological advances and diverse application domains. On the methodological side, this Special Issue includes studies on stimulus-dependent oculomotor responses, remote and video-based gaze detection for oculomotor assessment, event-camera imaging for robust pupil tracking, and machine-learning pipelines tailored to eye-movement analysis. On the application side, contributions address HMIs and operator monitoring, transportation (driver attention and distraction, situational awareness), aeronautics (documentation use and maintenance procedures), visual ergonomics and lens design, reading and educational assessment in children, and clinical or pre-clinical screening (e.g., early indicators relevant to learning difficulties). Collectively, they demonstrate how eye-movement data can inform design decisions, support objective evaluation, and unlock new forms of human-centered automation.

In curating this Special Issue, our aims were to (i) showcase instrumentation and signal-processing solutions that improve robustness and reproducibility in real-world contexts; (ii) highlight analytics and machine learning methods that transform raw gaze signals into actionable metrics; and (iii) present application case studies that connect those metrics to outcomes in safety, training, usability, and health. Across the contributions, common themes emerge: the need for principled handling of missing or noisy data; careful task and stimulus design to elicit diagnostic oculomotor patterns; integration of gaze with complementary modalities; and transparent evaluation protocols that enable cross-study comparison and deployment.

The remainder of this editorial provides a concise overview of each paper, emphasizing methodological innovations, datasets and experimental designs, and the implications of the reported findings for future research and practice. We conclude by outlining open challenges and opportunities for translating eye-tracking advances into reliable, ethical, and scalable solutions across domains.

2. An Overview of Published Articles

In the paper by Gomolka et al. (contribution 1), the authors investigate how deep neural networks can enhance the recording and interpretation of operator attention within advanced human–machine interfaces. Combining high-resolution eye-tracking with artificial-intelligence-based pattern recognition, they model cognitive states under varying workload conditions. Through controlled experiments, the team demonstrates that neural representations can capture subtle variations in fixation duration, saccadic dynamics, and visual focus stability corresponding to fluctuations in attention and fatigue. This integration of gaze analytics and deep learning reflects wider advances in real-time cognitive-state estimation [1,2,3,4,5,6,12,13]. The authors argue that such adaptive systems will play a key role in next-generation industrial automation, safety monitoring, and decision-support environments-domains in which intelligent gaze-based feedback can significantly improve operator performance and reliability [12,13,15,16].

The manuscript by Essig et al. (contribution 2) explores how optokinetic nystagmus (OKN) parameters depend on specific stimulus properties. Using a precision eye-tracking setup, the authors systematically vary contrast, motion direction, and texture to analyze corresponding changes in OKN amplitude, frequency, and latency. Their results reveal consistent modulation of reflexive eye movements by the physical structure of visual input, complementing recent studies on predictive oculomotor control and perceptual coherence [4,24,25]. The work contributes to both theoretical neuroscience and applied vision technology by providing methodological guidance for designing visual environments that elicit reliable involuntary eye responses-important for medical diagnostics and calibration protocols in oculomotor research.

In the study by Madlenak et al. (contribution 3), eye-tracking methods were applied to analyze the visual behavior of train drivers during real driving scenarios. The authors measured fixation points, saccade paths, and gaze dispersion to assess how attention is distributed among track signals, control elements, and the external environment. Their findings confirm that visual focus is predominantly directed toward safety-critical areas such as signals and speed indicators, while peripheral elements attract limited attention results consistent with earlier studies on operator attention and visual load dynamics in transport systems [29,30,31,32]. The discussion further considers how workload, environmental complexity, and route familiarity influence visual scanning strategies, aligning with cognitive-ergonomic models of driver monitoring and vigilance [33,34,35]. Overall, this contribution underscores the relevance of gaze analytics for enhancing driver training, cab-interface design, and real-time safety assessment in railway operations.

In the study by Lin et al. (contribution 4), the authors explored how stereoscopic 3D visual displays influence both ocular activity and motor performance. Using a controlled laboratory setting, participants were exposed to 2D and 3D stimuli while their saccade dynamics, fixation durations, and manual-response times were recorded. The results revealed measurable differences in gaze stability and reaction accuracy when interacting with 3D content, indicating an increased cognitive and oculomotor load. These findings are consistent with prior investigations into depth perception, visual fatigue, and sensorimotor coordination under immersive display conditions [29,30,31,32], which emphasize the complex interplay between binocular disparity and attention control. The authors discuss the ergonomic implications for prolonged use of 3D technologies in entertainment, education, and simulation, aligning with broader research on visual comfort optimization and adaptive display design [33,34,35]. Overall, the work contributes valuable insights into how advanced visual media affect human performance and perception.

In the work of Concepcion-Grande et al. (contribution 5), an innovative eye-tracking-based methodology is presented for assessing visual performance in users of progressive lens designs. The authors developed a quantitative evaluation framework linking fixation stability, gaze trajectory, and spatial attention metrics to optical lens parameters. Their controlled experiments demonstrate that the proposed system can identify subtle differences in visual adaptation and comfort between lens types, offering a reproducible, objective complement to conventional subjective testing. The study’s approach resonates with recent advances in visual behavior analytics and adaptive optics modeling [3,9,17,30], which emphasize the importance of integrating physiological data into optical-product validation. Moreover, the authors discuss how eye-movement signatures can inform ergonomic lens design and user-specific calibration [26,33,34], highlighting the broader applicability of gaze-based diagnostics in visual science research and ophthalmic engineering. This contribution provides a robust foundation for bridging laboratory-based measurements with practical applications in the optical wear industry.

In the study by Madlenak et al. (contribution 6), an integrated eye-tracking and A/B-testing approach is used to examine how outdoor advertising influences the visual attention and decision-making of car drivers in realistic driving scenarios. The authors collected gaze-distribution data from drivers exposed to various billboard designs differing in color, message complexity, and placement, and correlated these findings with recall and preference measures. The results indicate that vivid colors and concise message layouts elicit stronger fixation density and higher recall, whereas complex or text-heavy content tends to divert attention from the roadway. These outcomes correspond with earlier research on driver distraction and gaze-allocation behavior in dynamic visual environments [4,10,16,29,32], confirming the sensitivity of ocular metrics to perceptual load. The paper further discusses implications for traffic-safety policy, advertising ergonomics, and human-factors modeling [11,26,34], emphasizing the need to balance promotional effectiveness with attentional safety requirements in transport contexts.

In the study by Kang et al. (contribution 7), a novel event-camera imaging framework is evaluated as an emerging approach for remote pupil-tracking. Event cameras, or neuromorphic sensors, register only brightness changes at microsecond resolution, enabling high-speed, low-power acquisition of dynamic visual information. The authors demonstrate that this asynchronous imaging principle effectively captures rapid ocular movements and luminance variations even under variable lighting and head motion, outperforming conventional frame-based trackers. Their results are consistent with current advances in high-frequency gaze-sensing architectures and neuromorphic vision modeling [3,4,12,17,29], which seek to minimize latency and energy consumption. The paper highlights the potential of these sensors for automotive, mobile, and extended-reality systems, where motion blur and environmental variability often degrade optical performance. By integrating hardware efficiency with temporal precision, this contribution sets a technical foundation for next-generation real-time human–machine-interface and assistive-vision applications [11,33,34].

In their research, Kim et al. (contribution 8) analyzed how different reading modalities (such as digital and printed text) and text types influence the reading behavior of school-aged children through eye-tracking techniques. The study recorded fixation durations, saccade amplitudes, and regression frequencies to compare visual and cognitive processing across narrative and informational materials. Results revealed that digital formats elicited shorter fixations and more frequent regressions, indicating differences in comprehension strategies and visual strain. These findings correspond with prior work on reading fluency, visual ergonomics, and cognitive adaptation in digital literacy environments [15,17,25,30,32], highlighting developmental and media-related effects on reading efficiency. The authors further discuss pedagogical implications, suggesting that adaptive presentation and font optimization could mitigate ocular fatigue and support individualized learning. This contribution broadens the application of eye-movement analysis to educational technology and developmental psychology, offering a data-driven foundation for improving reading interfaces and cognitive-assessment tools [18,33,35].

In the study by Paris et al. (contribution 9), eye-tracking technology was applied to investigate how expertise level influences visual information processing during the use of aircraft maintenance documentation. The authors compared professional engineers and novice trainees as they performed diagnostic and procedural tasks, analyzing fixation density, gaze sequences, and information-search efficiency. Results show that experts exhibited shorter fixations and more structured visual patterns, indicating greater task familiarity and mental-model integration, whereas novices relied heavily on text scanning and repetitive verification. These outcomes align with established findings on expert–novice differences in cognitive workload, attention control, and domain-specific visual search [13,14,17,30,34], emphasizing how experience shapes perceptual organization. The paper concludes that eye-movement metrics can serve as objective indicators of technical proficiency and training effectiveness [7,18,35], supporting the development of adaptive instruction systems and digital maintenance interfaces optimized for safety and performance consistency.

In their work, Harpaz et al. (contribution 10) present an advanced video-based gaze-detection system designed to identify and quantify oculomotor abnormalities through precise, non-invasive measurement. The method combines high-resolution video acquisition with feature-tracking algorithms that automatically extract pupil position, velocity, and microsaccade parameters. The authors validated their approach using datasets from clinical and control populations, demonstrating high sensitivity in detecting subtle irregularities associated with neurological or visual motor disorders. The proposed framework aligns with ongoing developments in computer-vision-based diagnostics and medical eye-movement analytics [9,12,25,26,32], offering a practical balance between accessibility and diagnostic accuracy. Beyond its clinical relevance, the system contributes to broader applications in neuroergonomics, human–machine interaction, and rehabilitation monitoring [11,29,34], where reliable eye-movement assessment supports real-time evaluation of cognitive and sensorimotor functions. This study thus reinforces the potential of gaze-tracking technologies as robust, scalable tools for both research and applied health domains.

In their comprehensive literature review (contribution 11), Arias-Portela et al. analyze the role of eye-tracking metrics in assessing and enhancing drivers’ situational awareness within road-safety and human-factors research. The authors systematically synthesize prior empirical findings on fixation behavior, gaze entropy, and attention distribution under diverse environmental and cognitive-load conditions. Their review highlights that gaze-based indicators-such as fixation duration, saccade variability, and visual search efficiency can serve as reliable proxies for situational awareness, complementing subjective workload and reaction-time measures. The paper emphasizes methodological challenges related to dynamic-scene calibration, inter-subject variability, and cross-task comparability, echoing concerns raised in earlier studies on driver monitoring, vigilance assessment, and cognitive state estimation [4,8,10,29,32]. By outlining consistent analytical frameworks and integrating results from both simulator and on-road experiments, the authors establish a solid reference for future model development in adaptive driver-assistance systems and intelligent-transport interfaces [11,17,26,34].

In their experimental study, Gugerell et al. (contribution 12) investigated how visuospatial attention is distributed and managed during multitasking using the Multi-Attribute Task Battery II (MATB-II) framework. Participants performed concurrent flight-control, system-monitoring, and tracking tasks while their gaze patterns and response accuracy were analyzed to quantify attentional resource allocation. Results revealed that increased task complexity and switching frequency significantly altered fixation dwell times and scan-path organization, confirming that visuospatial attention dynamically adapts to workload and task prioritization. These findings correspond with prior work on attention management, visual cognitive control, and situational awareness in multitasking environments [3,8,10,17,29,32], emphasizing the interdependence between gaze dynamics and performance efficiency. The authors discuss implications for aerospace ergonomics and cognitive modeling, noting that eye-tracking data can serve as a real-time indicator for adaptive workload balancing in mission-critical systems. This contribution strengthens the empirical basis for designing attention-aware operator-support tools [11,26,34] in complex human–machine environments.

In their study, Nagy et al. (contribution 13) proposed and evaluated advanced area-of-interest (AOI) tracking techniques to detect and quantify visual distraction in simulated driving environments. The authors developed algorithms capable of dynamically segmenting gaze data into functional regions, such as road center, mirrors, dashboard, and peripheral zones, to capture momentary attention shifts and lapses. The method enables real-time identification of distraction events based on fixation duration and gaze-transition entropy, offering improved sensitivity over traditional AOI mapping. Results demonstrate that subtle deviations in visual behavior correlate with delayed response times and decreased situational awareness. These findings align with prior research on driver-attention modeling, visual load assessment, and gaze-based hazard prediction [6,10,16,17,29,32], highlighting the diagnostic potential of eye-movement analytics for intelligent transport safety. The study concludes that integrating AOI-driven metrics into driver-monitoring systems [7,11,26,34] can enhance the early detection of cognitive distraction and support adaptive vehicle-assistance technologies.

In their contribution, Zeslawska et al. (contribution 14) developed a hybrid diagnostic framework for early dyslexia detection that integrates eye-tracking metrics with long short-term memory (LSTM) neural networks. The approach captures detailed gaze trajectories and temporal features during children’s reading tasks, transforming them into sequential data suitable for deep-learning analysis. The trained LSTM model successfully distinguished dyslexic readers from typically developing peers with high classification accuracy, demonstrating the diagnostic power of combining physiological and computational methods. These results extend current research in machine-learning-assisted cognitive assessment and reading analytics [3,4,8,10,29,32], illustrating how data-driven techniques can complement traditional educational testing. The authors also discuss implications for early intervention, individualized therapy, and digital screening tools [11,17,26,33,34], emphasizing that integrating neural modeling with eye-tracking provides a scalable, non-invasive approach to developmental learning evaluation. This study concludes the collection by showcasing how modern AI- and vision-based analytics converge in applied cognitive science.

3. Conclusions

This Special Issue, “Eye-Tracking Technologies: Theory, Methods and Applications,” brings together a rich collection of studies that collectively illustrate the rapid maturation and diversification of modern eye-tracking research. Across the contributions, the papers reveal a strong convergence between technological innovation, cognitive modeling, and applied engineering. The works presented here address a wide spectrum of topics, ranging from fundamental investigations of visual physiology and oculomotor behavior to highly practical applications in transportation safety, industrial ergonomics, clinical diagnostics, and educational assessment. Taken together, they demonstrate that eye-tracking has evolved from a laboratory instrument for psychological inquiry into a multidisciplinary framework for understanding and augmenting human performance in complex environments [1,2,3,4,8,12].

Several contributions, including those by Gomolka et al., Kang et al., and Harpaz et al., show how deep-learning architectures, convolutional, recurrent, or attention-based can extract meaningful patterns from gaze dynamics in real time. These developments move eye-tracking beyond descriptive statistics toward predictive and adaptive modeling, mirroring advances in transformer-based architectures, semi-supervised learning, and neuromorphic sensing [1,2,3,4,5,6,7,12,13]. The resulting systems can automatically recognize cognitive states, detect anomalies, and infer task intent. Such capabilities open new possibilities for human–machine interaction, enabling interfaces that dynamically respond to the user’s level of attention, fatigue, or comprehension [12,13,14,15,16,17,18]. In parallel, improvements in computational efficiency and the use of consumer-grade imaging hardware lower the barriers to deployment, expanding the reach of gaze-based analytics beyond specialized laboratories.

Another recurrent topic concerns visual attention and situational awareness in transport and control domains. Papers by Madlenak et al., Nagy et al., and Arias-Portela et al. collectively demonstrate the practical value of eye-tracking for monitoring operator performance in driving, railway, and pedestrian-safety contexts. These studies show how fixation sequences, scanning strategies, and AOI distributions correspond to safety-critical behaviors [14,15,16,17,18]. The insights gained have direct implications for the design of dashboards, driver-assistance systems, and training programs, contributing to a deeper understanding of human reliability in dynamic multitasking environments. At the same time, the systematic review by Arias-Portela et al. underscores the importance of methodological consistency and multimodal data fusion to ensure comparability and reproducibility across driver-monitoring studies [12,13,14,15,16,17,18].

A third important line of research relates to visual ergonomics, perceptual comfort, and human factors. The papers by Essig et al., Lin et al., and Concepcion-Grande et al. explore how specific visual stimulus properties contrast, motion, stereoscopic depth, and calibration strategy-shape ocular behavior and influence comfort, accuracy, and fatigue. These results demonstrate that quantitative gaze analysis provides objective, reproducible indicators of user experience, complementing subjective ratings and self-reports. Such evidence-based assessment supports innovation in optical design, extended-reality visualization, and human–machine interfaces [4,19,20,21,22,26]. More broadly, it illustrates how eye-tracking contributes to the ergonomics of perception, offering measurable insight into how humans adapt to increasingly immersive digital environments.

Equally prominent is the use of eye-tracking in education, training, and clinical diagnostics. The studies by Kim et al. and Gomolka et al. highlight how gaze metrics can identify early indicators of dyslexia and other reading-related difficulties, while Harpaz et al. and Madlenak et al. demonstrate how oculomotor and behavioral data can monitor cognitive workload, fatigue, and attention dynamics [7,8,9,10,11,13,16]. These examples show the transformative potential of gaze analytics in supporting individualized learning, adaptive training, and accessible healthcare diagnostics. The integration of deep-learning methods with biomedical signal processing [7,8,9,11,23] points toward a future where eye-tracking acts as a non-invasive window into neurological and cognitive function, bridging neuroscience, education, and clinical practice.

Beyond these thematic clusters, several cross-cutting methodological trends emerge. A clear movement toward real-world data collection using wearable, mobile, or vehicle-mounted sensors enables authentic behavioral monitoring in ecological contexts [4,14,15,16,17]. Simultaneously, the field is transitioning from purely descriptive analyses toward model-based frameworks capable of linking gaze dynamics to decision processes, workload, and affective states [12,13,16,21,22,26]. Contributions throughout the issue emphasize the need for standardized data formats, transparent evaluation protocols, and ethical handling of biometric information. Addressing these challenges will be vital for ensuring reproducibility, privacy, and the responsible integration of eye-tracking into daily life.

Looking forward, future eye-tracking systems are expected to benefit from continued advances in sensor miniaturization, neuromorphic imaging, and multimodal fusion with physiological and behavioral data streams [3,4,7,11]. These innovations will support new forms of intelligent, context-aware systems capable of continuous interpretation of user intent and affect. At the same time, the democratization of hardware and software calls for heightened attention to data governance, interpretability of machine-learning models, and safeguards against misuse. Sustained interdisciplinary collaboration among engineers, neuroscientists, psychologists, clinicians, and educators remains essential for translating technological progress into socially beneficial outcomes.

In conclusion, the contributions assembled in this Special Issue demonstrate both the maturity and vitality of contemporary eye-tracking research. They collectively point toward a future in which gaze-based sensing forms a cornerstone of human-centered design, adaptive automation, and cognitive-health assessment. By bridging theoretical insight with applied innovation, this volume highlights the enduring relevance of eye-tracking as a means of exploring the interface between perception, cognition, and intelligent technology.