Publications

Explore a collection of publications and projects, from diverse fields, that cite Pupil Labs and use Pupil Labs eye tracking hardware and software in their research.

Filters

Year

101
64
55
47
40
49
25
9

Product

323
43
16
8

Fields

80
54
44
39
35
33
27
27
25
24
23
20
18
17
13
11
11
10
10
9
6
6
6
6
6
5
5
5
5
5
5
5
4
4
4
4
4
3
3
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0-0 of 390 publications
What Are You Looking At? Using Eye Tracking Glasses to Monitor Toddler Attention in Natural Learning Situations.
2021
Developmental Psychology
Nicole Altvater-Mackensen
Eye movements are taken to reflect cognitive processes. Because eye tracking is non-invasive and does not require an overt behavioural response, it is a suitable method to use with infants and toddlers. Most eye tracking studies to date use screen-based remote systems which allow high experimental control but limit the possibilities to test participants in more naturalistic, interactive settings. In this paper, we present a prototype of mobile eye tracking glasses suitable for toddlers and describe its use in a set-up that allows for automatic gaze coding. We present the components and set-up of the eye tracking glasses, as well as the basic principles and routines of data acquisition and analysis. Data from a pilot study testing 2- to 5-year-olds in a shared reading interaction and a preferential looking task is used to illustrate data quality and potential pitfalls in data collection.
An implicit representation of stimulus ambiguity in pupil size
2021
audio-visual, pupilometry, cognition
Graves, J. E., Egré, P., Pressnitzer, D., & de Gardelle, V.
Procedings of the national academy of sciences
To guide behavior, perceptual systems must operate on intrinsically ambiguous sensory input. Observers are usually able to acknowledge the uncertainty of their perception, but in some cases, they critically fail to do so. Here, we show that a physiological correlate of ambiguity can be found in pupil dilation even when the observer is not aware of such ambiguity. We used a well-known auditory ambiguous stimulus, known as the tritone paradox, which can induce the perception of an upward or downward pitch shift within the same individual. In two experiments, behavioral responses showed that listeners could not explicitly access the ambiguity in this stimulus, even though their responses varied from trial to trial. However, pupil dilation was larger for the more ambiguous cases. The ambiguity of the stimulus for each listener was indexed by the entropy of behavioral responses, and this entropy was also a significant predictor of pupil size. In particular, entropy explained additional variation in pupil size independent of the explicit judgment of confidence in the specific situation that we investigated, in which the two measures were decoupled. Our data thus suggest that stimulus ambiguity is implicitly represented in the brain even without explicit awareness of this ambiguity.
Social Virtual Reality: Implementing Non-verbal Cues in Remote Synchronous Communication
2021
VR
Kasapakis, V., Dzardanova, E., Nikolakopoulou, V., Vosinakis, S., Xenakis, I., & Gavalas, D.
International Conference on Virtual Reality and Mixed Reality
Social Virtual Reality (SVR) platforms allow remote, synchronous interaction and communication between individuals immersed in shared virtual worlds. Such platforms commonly implement full-body motion and real-time voice communication, but often lack complete non-verbal cues support. This work presents the development process and preliminary usability evaluation results of an SVR platform, incorporating non-verbal cues such as finger motion, gaze direction, and facial expressions, while allowing inter-communication between remotely located interlocutors.
Binocular vision and the control of foot placement during walking in natural terrain.
2021
Navigation, Walking, Motor Control
Bonnen, K., Matthis, J. S., Gibaldi, A., Banks, M. S., Levi, D. M., & Hayhoe, M.
Scientific Reports (Nature)
Coordination between visual and motor processes is critical for the selection of stable footholds when walking in uneven terrains. While recent work (Matthis et al. in Curr Biol 8(28):1224–1233, 2018) demonstrates a tight link between gaze (visual) and gait (motor), it remains unclear which aspects of visual information play a role in this visuomotor control loop, and how the loss of this information affects that relationship. Here we examine the role of binocular information in the visuomotor control of walking over complex terrain. We recorded eye and body movements while normally-sighted participants walked over terrains of varying difficulty, with intact vision or with vision in one eye blurred to disrupt binocular vision. Gaze strategy was highly sensitive to the complexity of the terrain, with more fixations dedicated to foothold selection as the terrain became more difficult. The primary effect of increased sensory uncertainty due to disrupted binocular vision was a small bias in gaze towards closer footholds, indicating greater pressure on the visuomotor control process. Participants with binocular vision losses due to developmental disorders (i.e., amblyopia, strabismus), who have had the opportunity to develop alternative strategies, also biased their gaze towards closer footholds. Across all participants, we observed a relationship between an individual’s typical level of binocular visual function and the degree to which gaze is shifted toward the body. Thus the gaze–gait relationship is sensitive to the level of sensory uncertainty, and deficits in binocular visual function (whether transient or long-standing) have systematic effects on gaze strategy in complex terrains. We conclude that binocular vision provides useful information for locating footholds during locomotion. Furthermore, we have demonstrated that combined eye/body tracking in natural environments can be used to provide a more detailed understanding of the impact of a type of vision loss on the visuomotor control process of walking, a vital everyday task.
Gaze-Adaptive Subtitles Considering the Balance among Vertical/Horizontal and Depth of Eye Movement.
2021
AR
Shimizu, Y., Ohnishi, A., Terada, T., & Tsukamoto, M.
IEEE International Symposium on Mixed and Augmented Reality Adjunct
Subtitles (captions displayed on the screen) are important in 3D content, such as virtual reality (VR) and 3D movies, to help users understand the content. However, an optimal displaying method and framework for subtitles have not been established for 3D content because 3D has a depth factor. To determine how to place text in 3D content, we propose four methods of moving subtitles dynamically considering the balance between the vertical/horizontal and depth of gaze shift. These methods are used to reduce the difference in depth or distance between the gaze position and subtitles. Additionally, we evaluate the readability of the text and participants’ fatigue. The results show that aligning the text horizontally and vertically to eye movements improves visibility and readability. It is also shown that the eyestrain is related to the distance between the object and subtitles. This evaluation provides basic knowledge for presenting text in 3D content.
Human gaze-aware attentive object detection for ambient intelligence.
2021
HCI, AR
Cho, D. Y., & Kang, M. K.
Engineering Applications of Artificial Intelligence
Understanding human behavior and the surrounding environment is essential for realizing ambient intelligence (AmI), for which eye gaze and object information are reliable cues. In this study, the authors propose a novel human gaze-aware attentive object detection framework as an elemental technology for AmI. The proposed framework detects users’ attentive objects and shows more precise and robust performance against object-scale variations. A novel Adaptive-3D-Region-of-Interest (Ada-3D-RoI) scheme is designed as a front-end module, and scalable detection network structures are proposed to maximize cost-efficiency. The experiments show that the detection rate is improved up to 97.6% on small objects (14.1% on average), and it is selectively tunable with a tradeoff between accuracy and computational complexity. In addition, the qualitative results demonstrate that the proposed framework detects a user’s single object-of-interest only, even when the target object is occluded or extremely small. Complementary matters for follow-up study are presented as suggestions to extend the results of the proposed framework to further practical AmI applications. This study will help develop advanced AmI applications that demand a higher-level understanding of scene context and human behavior such as human–robot symbiosis, remote-/autonomous control, and augmented/mixed reality.
Integrating Mobile Eye-Tracking and VSLAM for Recording Spatial Gaze in Works of Art and Architecture
2021
Eye Tracking, Attention, Visualization, Architecture
Eugene Han
Technology|Architecture + Design
This paper proposes a method for spatial eye-tracking for use with three-dimensional objects and environments. By integrating mobile eye-tracking (MET) with Visual Simultaneous Localization and Mapping (VSLAM) technologies, the study provides an unobtrusive technique for capturing an individual’s gaze across an open space and through an unprescribed viewing path. The included proof-of-concept is tested against three scales, from a large sculpture to a set of towering cement kilns and an exterior building passage. Demonstrations show that the integration of MET and VSLAM provides a useful tool for testing scenarios without predefined viewing conditions and allowing insight into how others view works of art and architecture.
CorrNet: Fine-Grained Emotion Recognition for Video Watching Using Wearable Physiological Sensors
2021
Psychology
Tianyi Zhang, Abdallah El Ali, Chen Wang, Alan Hanjalic and Pablo Cesar
MDPI: Sensors
Recognizing user emotions while they watch short-form videos anytime and anywhere is essential for facilitating video content customization and personalization. However, most works either classify a single emotion per video stimuli, or are restricted to static, desktop environments. To address this, we propose a correlation-based emotion recognition algorithm (CorrNet) to recognize the valence and arousal (V-A) of each instance (fine-grained segment of signals) using only wearable, physiological signals (e.g., electrodermal activity, heart rate). CorrNet takes advantage of features both inside each instance (intra-modality features) and between different instances for the same video stimuli (correlation-based features). We first test our approach on an indoor-desktop affect dataset (CASE), and thereafter on an outdoor-mobile affect dataset (MERCA) which we collected using a smart wristband and wearable eyetracker. Results show that for subject-independent binary classification (high-low), CorrNet yields promising recognition accuracies: 76.37% and 74.03% for V-A on CASE, and 70.29% and 68.15% for V-A on MERCA. Our findings show: (1) instance segment lengths between 1–4 s result in highest recognition accuracies (2) accuracies between laboratory-grade and wearable sensors are comparable, even under low sampling rates (≤64 Hz) (3) large amounts of neutral V-A labels, an artifact of continuous affect annotation, result in varied recognition performance.
A Study on Evacuation Behavior in Physical and Virtual Reality Experiments
2021
VR, Eye Tracking
Silvia Arias, Axel Mossberg, Daniel Nilsson & Jonathan Wahlqvist
Fire Technology
Comparing results obtained in Virtual Reality to those obtained in physical experiments is key for validation of Virtual Reality as a research method in the field of Human Behavior in Fire. A series of experiments based on similar evacuation scenarios in a high-rise building with evacuation elevators was conducted. The experiments consisted of a physical experiment in a building, and two Virtual Reality experiments in a virtual representation of the same building: one using a Cave Automatic Virtual Environment (CAVE), and one using a head-mounted display (HMD). The data obtained in the HMD experiment is compared to data obtained in the CAVE and physical experiment. The three datasets were compared in terms of pre-evacuation time, noticing escape routes, walking paths, exit choice, waiting times for the elevators and eye-tracking data related to emergency signage. The HMD experiment was able to reproduce the data obtained in the physical experiment in terms of pre-evacuation time and exit choice, but there were large differences with the results from the CAVE experiment. Possible factors affecting the data produced using Virtual Reality are identified, such as spatial orientation and movement in the virtual environment.
Profiles of naturalistic attentional trajectories associated with internalizing behaviors in school-age children: A mobile eye tracking study Kelley
2021
Behavioral Inhibition, Attention
Kelley E. Gunther, Xiaoxue Fu, Leigha MacNeill, Alicia Vallorani, Briana Ermanni, & Koraly Pérez-Edgar
Naturalistic Attentional Trajectories
The temperament profile Behavioral Inhibition (BI) is a strong predictor of internalizing behavior in childhood. Patterns of attention towards or away from threat are a commonality of both BI and internalizing behaviors. Attention biases are traditionally measured with computer tasks presenting affective stimuli, which can lack ecological validity. Recent studies suggest that naturalistic visual attention need not mirror findings from computer tasks, and, more specifically, children high in BI may attend less to threats in naturalistic tasks. Here, we characterized latent trajectories of naturalistic visual attention over time to a female stranger, measured with mobile eye tracking, among kindergarteners oversampled for BI. Group-based trajectory modeling (GBTM) revealed two latent trajectories: 1) high initial orienting to the stranger, gradual decay, and recovery, and 2) low initial orienting and continued avoidance. Higher probability of membership to the “avoidant” group was linked to greater report of internalizing behaviors. We demonstrate the efficacy of mobile eye tracking in quantifying naturalistic patterns of visual attention to social novelty, as well as the importance of naturalistic measures of attention in characterizing socioemotional risk factors.