Gaze Angle Research Articles

Ultrasound for Gaze Estimation-A Modeling and Empirical Study.

Most eye tracking methods are light-based. As such, they can suffer from ambient light changes when used outdoors, especially for use cases where eye trackers are embedded in Augmented Reality glasses. It has been recently suggested that ultrasound could provide a low power, fast, light-insensitive alternative to camera-based sensors for eye tracking. Here, we report on our work on modeling ultrasound sensor integration into a glasses form factor AR device to evaluate the feasibility of estimating eye-gaze in various configurations. Next, we designed a benchtop experimental setup to collect empirical data on time of flight and amplitude signals for reflected ultrasound waves for a range of gaze angles of a model eye. We used this data as input for a low-complexity gradient-boosted tree machine learning regression model and demonstrate that we can effectively estimate gaze (gaze RMSE error of 0.965 ± 0.178 degrees with an adjusted score of 90.2 ± 4.6).

The Cone of Direct Gaze: A Stable Trait.

Direct eye gaze is a potent stimulus in social interactions and is often associated with interest and approach orientation. Yet, there is remarkable variability in the range of gaze lines that people accept as being direct. A measure that is frequently used to quantify the range of gaze angles within which an observer assumes mutual gaze is the cone of direct gaze (CoDG). While individual differences in CoDG have often been examined, studies that systematically investigate the stability of an observers' CoDG over time are scarce. In two experiments, we measured the CoDG using an established paradigm and repeated the measurement after 5 min and/or after 1 week. We found high inter-individual variation, but high agreement within participants (ICCs between 0.649 and 0.855). We conclude that the CoDG can be seen as a rather stable measure, much like a personality trait.

Iris-ocular-periocular: toward more accurate biometrics for off-angle images

Iris is one of the most well-known biometrics; it is a nonintrusive and contactless authentication technique with high accuracy, enhanced security, and unique distinctiveness. However, its dependence on image quality and its frontal image acquisition requirement limit its recognition performance and hinder its potential use in standoff applications. Standoff biometric systems require a less controlled environment than traditional systems, so their captured images will likely be nonideal, including off-angle. We present convolutional neural network (CNN)-based deep learning frameworks to improve the recognition performance of iris, ocular, and periocular biometric modalities for off-angle images. Our contribution is fourfold: first, the performances of popular AlexNet, GoogLeNet, and ResNet50 architectures are presented for off-angle biometrics. Second, we study the effect of the gaze angle difference between training and testing images on iris, ocular, and periocular recognitions. Third, we investigate the network behavior for untrained gaze angles and the information fusion capability of CNN networks at multiple off-angle images. Finally, deep learning-based results are compared with a traditional iris recognition algorithm using the gallery approach. Our results with off-angle images ranging from −50 deg to 50 deg in gaze angle show that the proposed methods improve the recognition performance of iris, ocular, and periocular recognition.

A New Clinical Method for Visual Function Evaluation Including Estimation and Measurement of Therapeutic Effectiveness: Lessons From Infantile Nystagmus Research.

To translate infantile nystagmus system (INS) research into easily understood, clinically relevant terminology and suggest modifications to research and clinical testing, data and clinical interpretation, and therapeutic choices and evaluation. A clinical method is presented using only three best-corrected visual acuity measurements of patients with INS, whereby (1) a measure of the quality of visual acuity across the visual field is possible; (2) pre-therapy estimates of post-therapy improvements in peak acuity and the high-acuity range of gaze angles are possible; and (3) more realistic visual function outcome measures of therapy are available to the practitioner. The application of the high-acuity field quality spreadsheet to the analyses of patients with INS (before and after therapy) results in a quantitative measure of visual function based on three visual acuity measurements. The clinician can now duplicate adequate functional visual acuity descriptions in patients with INS along with their pre-therapy estimates and outcome measures. Previously, these have only been available to researchers or the rare clinicians who have access to both eye movement data and the expanded nystagmus acuity function analysis of INS waveforms. [J Pediatr Ophthalmol Strabismus. 2021;58(3):188-195.].

Visual Sequelae of Computer Vision Syndrome: A Cross-Sectional Case-Control Study.

Purpose To assess the visual, ocular, extraocular, and multifocal electroretinography (mfERG) outcomes of computer vision syndrome (CVS) among medical students. Methods This study was designed as a cross-sectional case-control study that included 733 medical students. All students completed a specially designed and validated CVS questionnaire survey (CVS-F3). Students from the control (No-CVS) and CVS groups underwent comprehensive ophthalmic examinations including the mfERG examinations. Our main outcome measures included uncorrected and corrected distance visual acuity (UDVA and CDVA, resp.) measurements, subjective and cycloplegic refractions, slit-lamp examination, intraocular pressure measurement, pupillary reflexes tests, ocular movements' tests, dry eye disease tests, and fundus and mfERG examinations. Results The CVS-F3 identified that 87.9% of students had complaints that might be related to CVS. We documented a 76% prevalence rate in students undergoing an ophthalmologic exam. The most common ocular and extraocular complaints included visual blur and headache (40.9% and 46.8%, resp.). Statistical logistic and linear regression analyses showed that refractive errors, prolonged screen-hours, close eye-screen distance, improper gaze angle, poor screen-resolution, and screen-glare were risk factors for developing CVS and influencing its severity. In the mfERG subgroup, 42.5% demonstrated reduced amplitudes of mfERG rings and quadrants, indicating reduced foveal responses. Conclusion Surveys cannot yield an accurate CVS prevalence. However, they help to identify subjects at risk who should be comprehensively assessed to confirm or exclude CVS diagnosis. Smartphone misuse primarily caused CVS among users. Our mfERG findings might be a sign of potential CVS visual sequelae; however, future studies are warranted. Clinicians need to understand these sequelae to appropriately identify and treat CVS.

Evaluation of the Effect of the Restoration of Lost Vertical Dimension with Permissible Splint on Posture using Photogrammetry and S-EMG - An In-Vivo Study

Introduction: Occlusion is a multidimensional phenomenon of which vertical dimension is an important factor. Maintenance of occlusal harmony along vertical dimension helps in better prognosis of the treatment performed. Along with the stomatognathic system, it has also been found to be beneficial for overall health of the postural muscles. Therefore, co-relation between vertical dimension and postural muscles analysis is significant. Aim/Objectives: To evaluate the effect of the restoration of lost vertical dimension with a permissible splint on posture using photogrammetry and S-EMG. Methods and Material: Fifteen patients were selected according to the inclusion and exclusion criteria for the study. Vertical dimension of rest and occlusion were measured using Niswonger’s technique. Posture was evaluated using photogrammetry and S-EMG. For photogrammetry, CVA angle and Gaze angle were measured using photographs on Kinovea software. The posture was evaluated by measuring the activity of postural muscles, namely Sternocleidomastoid [SCM], upper trapezius and spinous erectae using S-EMG. Permissible splints were constructed at restored vertical relation and given to the patient after occlusal equilibration. 30 days post-restoration, the posture was re-evaluated using same methods. Results: Difference between right and left side reduced from pre to post but was not statistically significant. Relation was found only between right upper trapezius and right spinous erectae and left upper trapezius and left SCM. Conclusion: Posture is affected by a number of factors, one of which happens to be occlusion. This study proved that occlusion does have an effect on body posture and an improvement in occlusion will only aid in improving body posture

Low-Profile Induced-Voltage Distance Ranger for Smart Contact Lenses.

In this paper, we present a novel, low-profile, scleral-coil based, distance ranging system which is suitable for smart, accommodating contact lenses. We measure the induced emf between a set of four thin semi-circular coils patterned on flexible Kapton substrates that conform to the eyes' sclera. This induced emf is a function of eye gaze angles. The system then determines the distance from the eyes to the desired object via the triangulation of these eye gaze angles Results: Experiments on eyeball simulated tissue gels indicate an accurate prediction of object distance in the 0.1-15 D range with a 0.15 D RMS error and object direction in the -15 to 15-degree arc with 0.4-degree RMS error, respectively. The energy required was determined to be as low as 20 μJ per range reading. Experimental data shows that our proposed new method of eye-tracking and distance ranging system can accurately predict eye-gaze angles and object-distance, whilst using only 20 μJ per range reading. The high-accuracy, low-profile and reduced energy requirements of the proposed eye-tracking technique, make it suitable for applications in the vast field of adaptive optics such as smart contact lenses and other low-power vision corrective applications.

Dynamic causal modeling of eye gaze processing in schizophrenia

BackgroundAbnormal eye gaze perception is related to symptoms and social functioning in schizophrenia. However, little is known about the brain network mechanisms underlying these abnormalities. Here, we employed dynamic causal modeling (DCM) of fMRI data to discover aberrant effective connectivity within networks associated with eye gaze processing in schizophrenia. MethodsTwenty-seven patients (schizophrenia/schizoaffective disorder, SZ) and 22 healthy controls (HC) completed an eye gaze processing task during fMRI. Participants viewed faces with different gaze angles and performed explicit gaze discrimination (Gaze: “Looking at you?” yes/no) or implicit gaze processing (Gender: “male or female?”). Four brain regions, the secondary visual cortex (Vis), posterior superior temporal sulcus (pSTS), inferior parietal lobule (IPL), and posterior medial frontal cortex (pMFC) were identified as nodes for subsequent DCM analysis. ResultsSZ and HC showed similar generative model structure, but SZ showed altered connectivity for specific self-connections, inter-regional connections during all gaze processing (reduced excitatory bottom-up and enhanced inhibitory top-down connections), and modulation by explicit gaze discrimination (increased frontal inhibition of visual cortex). Altered effective connectivity was significantly associated with poorer social cognition and functioning. ConclusionsGeneral gaze processing in SZ is associated with distributed cortical dysfunctions and bidirectional connectivity between regions, while explicit gaze discrimination involves predominantly top-down abnormalities in the visual system. These results suggest plausible neural mechanisms underpinning gaze processing deficits and may serve as bio-markers for intervention.

Gaze behavior during urban cycling: Effects of subjective risk perception and vista space properties

In this research, we investigate how gaze behavior during urban cycling is affected by subjective risk perception and the local spatial configuration. For this purpose, participants cycled at five defined test locations while wearing a mobile eye tracking system. Next to the fixation duration, we extracted sight vector lengths (i.e., the distance between the current body location and the gaze location) and gaze angles (i.e., the degree to which the gaze deviates from the current travel direction). Concerning risk perception, participants provided a hazard estimate for each test location, as well as information about specific areas they experienced as dangerous. This information was aggregated into a local hazard level across all participants. Finally, the vista space (i.e., the space visible from the current position during a fixation) was quantified via isovist analysis. The results imply that spatially open locations, extending about equally into all directions, increase the level of perceived risk. Depending on the local spatial properties, participants also gaze into different directions, rather than further ahead and into their travel direction. Higher cycling experience and greater familiarity with a location enabled a more foresighted and focused gaze behavior. Our findings underline the importance to provide cyclists not only with enough space and with unobstructed views, but also to limit the space they are required to distribute their attention to.

Key Clinical Concepts from Multi-Institutional Experience with Dedicated Fixed Low-Energy Proton Ocular Beamlines

Proton beam treatment for ocular tumors has been well established with dedicated eye beamlines internationally. Given the emergence of universal, high-energy, non-fixed beamline options, we share key practices for uveal melanoma (UM) treatment planning and delivery. We focus particularly on practical clinical and physics concepts to minimize important ocular side effects. Practices were studied from two well-established international dedicated proton ocular institutions with very longstanding facilities, both using fixed, low-energy, ocular beamlines to treat eye tumors only. Since the 1970s and 1980s until now, the respective cyclotrons produce 67.5 and 60.0 MeV proton beams (range up to 3 cm). In total 5927 ocular patients, 93% with UM, have been treated. Standard UM dose is 56 Gray Equivalent (GyE) and 57.2 GyE, respectively, delivered in 4 fractions. Planning and delivery procedures were analyzed as a bi-institutional effort to communicate key concepts to newer developing proton centers. Normal eye tissue sparing is related to critical beam characteristics provided in the setting of low-energy dedicated eye beamlines. Both centers offer a very sharp dose fall off (distal ∼1 mm and lateral ∼1.1-1.6 mm), high dose homogeneity, excellent range precision and short treatment time (∼0.5-2 minutes). Planned doses to critical ocular structures are independent predictors of ocular complications such as vision loss and neovascular glaucoma, including the following parameters: 28 GyE to macula [P<.0001], optic nerve [P<0.0004], lens [P<.0001], and ciliary body [P<.0001]. A systematic tumor/critical structure dose evaluation is used to optimize treatment parameters, i.e. lateral margin, distal range, gaze angle, and aperture shape. New high-energy non-dedicated beam designs, which degrade energy to deliver ocular treatment, may require significant beam adjustment and treatment planning procedures to achieve adequate characteristics. Regarding UM dose, a common regimen is 60 GyE in 4 fractions globally. Unrandomized early retrospective data shows lower local tumor control with 48 GyE (P = 0.02). Critical structures including the retina, lacrimal gland, cornea, tear ducts, lids, bony orbit, and limbal stem cells are methodically evaluated to minimize side effects. Eyelid toxicity is reduced by institutional retraction techniques or treatment through closed lids. Experienced international specialty proton ocular centers with dedicated, low-energy, fixed eyelines, provide important practical concepts for consideration by developing centers to optimize high-energy universal beamline designs for eye treatments. As new centers with different beam designs emerge globally, key clinical and treatment planning/delivery practices will need to be applied, particularly to reduce potential normal tissue toxicity for ocular patients.

Measuring Driver Perception: Combining Eye-Tracking and Automated Road Scene Perception.

ObjectiveTo investigate how well gaze behavior can indicate driver awareness of individual road users when related to the vehicle’s road scene perception.BackgroundAn appropriate method is required to identify how driver gaze reveals awareness of other road users.MethodWe developed a recognition-based method for labeling of driver situation awareness (SA) in a vehicle with road-scene perception and eye tracking. Thirteen drivers performed 91 left turns on complex urban intersections and identified images of encountered road users among distractor images.ResultsDrivers fixated within 2° for 72.8% of relevant and 27.8% of irrelevant road users and were able to recognize 36.1% of the relevant and 19.4% of irrelevant road users one min after leaving the intersection. Gaze behavior could predict road user relevance but not the outcome of the recognition task. Unexpectedly, 18% of road users observed beyond 10° were recognized.ConclusionsDespite suboptimal psychometric properties leading to low recognition rates, our recognition task could identify awareness of individual road users during left turn maneuvers. Perception occurred at gaze angles well beyond 2°, which means that fixation locations are insufficient for awareness monitoring.ApplicationFindings can be used in driver attention and awareness modelling, and design of gaze-based driver support systems.

Dosimetry modeling of focused kV x‐ray radiotherapy for wet age‐related macular degeneration

Wet (neovascular) age-related macular degeneration (AMD) is the leading cause of blindness in the United States. The mainstay treatment requires monthly intravitreal injection of anti-vascular endothelial growth factor (anti-VEGF) drugs, associated with multiple visits, high cost, and the risk of procedural injury and infection. Anti-VEGF drugs inhibit the formation of neovasculature but do not directly attack it. Radiotherapy can destroy neovasculature and potentially also inhibit wet-AMD associated inflammation and fibrosis not addressed by VEGF inhibitors. However, the current collimation-based radiotherapy device uses fixed 4mm beams, which are prone to overtreat or undertreat the choroidal neovascularization (CNV) lesions because of their various sizes and shapes. This simulation study evaluates personalized conformal treatment with focused kV radiation using cutting-edge polycapillary x-ray optics. Simulation of the polycapillary optics was achieved via Monte Carlo (MC)-based three-dimensional (3D) geometric ray tracing. Phase-space files modeling the focused photons were generated. The method was previously verified by phantom measurements. The ultrasmall ~0.2mm beam focal spot perpendicular to the beam direction enables spatially fractionated grid therapy, which has been shown to preferentially damage abnormal neovascular blood vessels vs normal ones. Geant4-based MC simulations of scanning while rotating beam delivery were performed to conformally treat three clinical cases of large, medium, and small CNV lesions with regular and grid deliveries. Dose delivery uncertainties due to positioning errors were analyzed, including ±0.75mm displacement in the three orthogonal directions and ±5° vertical/horizontal rotation of the eyeball. The simulated CNV treatments by 60-kVp focused x-ray beams show highly conformal delivery of dose to the lesion plus margin (0.75mm) with sharp dose fall-offs and controllable spatial modulation patterns. The 90%-10% isodose penumbra is <0.5mm. With a prescription dose of 16Gy to the lesions, the critical structure doses are well below the tolerance. The average CNV dose varies within 10% (mostly within 4%) due to 0.75-mm linear displacements and 5-degree gaze angle rotation of the eyeball. Focused kV technique allows personalized treatment of CNV lesions and reduces unwanted radiation to adjacent healthy tissue. The simulated dose distribution is superior to currently available techniques.

A Novel Camera-Free Eye Tracking Sensor for Augmented Reality Based on Laser Scanning

Next generation AR glasses require a highly integrated, high-resolution near-eye display technique such as focus-free retinal projection to enhance usability. Combined with low-power eye-tracking, such glasses enable better user experience and performance. We propose a novel and robust low-power eye-tracking sensor for integration into retinal projection systems. In our approach, a MEMS micro mirror scans an IR laser beam over the eye region and the scattered light is received by a photodiode. The advantages of our approach over typical VOG systems are its high integration capability and low-power consumption, which qualify our approach for next generation AR glasses. Here, we present a mathematical framework to estimate the achievable gaze angle resolution of our approach. We further show the viability of the proposed eye-tracking sensor based on a laboratory setup and discuss power consumption and gaze angle resolution compared to typical eye-tracking techniques.

Fractal analysis of retinal vasculature in normal subjects on ultra-wide field fluorescein angiography.

To evaluate the fractal feature of the retinal vasculature of normal eyes on a stereographic projected and montaged ultra-wide field (UWF) fluorescein angiography (FA). Prospective, observational, cross-sectional study. Totally 59 eyes of 31 normal subjects were imaged using the Optos 200Tx. Images obtained at different gaze angles stereographically projected and montaged. The early-phase UWF FA frames were processed to segment the retinal vasculature and the results were exported as binary masks. The fractal dimension (FD) was calculated using the box-counting method. The global FD for the entire retina was 1.6±0.04, with no difference between males and females (1.59±0.04 vs 1.61±0.04, P=0.084) or between right and left eyes (1.6±0.04 vs 1.6±0.05, P=0.61). FD was non-uniformly distributed among four quadrants (P<0.001) and decreased as the distance from the fovea increased (P<0.001). A negative association was observed between FD and age (R=-0.37, P=0.006), and this relationship was observed in the posterior and mid-peripheral retina (P<0.05) but absent in far-periphery (P>0.05). Fractal geometry is non-uniformly distributed across the retina in normal eyes and decreases from the fovea to the far-periphery. Subjects with an older age tend to have a smaller FD, however, the FD in the far-periphery does not appear to be influenced by age.

Evaluation of mental workload during automobile driving using one-class support vector machine with eye movement data

The aim of this study is to investigate the usefulness of the anomaly detection method by one-class support vector machine (OCSVM) for the evaluation of mental workload (MWL) during automobile driving. Twelve students (six males and six females) participated. The participants performed driving tasks with a driving simulator (DS) and the N-back task that was used to control their MWL. The N-back task had five difficulty levels from “none” to “3-back.” Eye and head movements were measured during the DS driving. Results showed that the standard deviation (SD) of the gaze angle, SD of eyeball rotation angle, share rate of head movement, and blink frequency had significant correlations with the task difficulty. The decision boundary of OCSVM could detect 95% of high MWL state (i.e., “3-back” state). In addition, the absolute value of the distance from the decision boundary increased with the task difficulty from “0-back” to “3-back.”

Impact of Fixation Direction on Dynamic Accommodative Response

Purpose : The purpose of this study was to measure accommodative response when looking at different direction such as downgaze and sidegaze either right or left. Methods : This study was performed with 13 adult males and females(22.40±0.41 years old) no subject had ophthalmic surgery or disease and placed the targets at the primary, 10° and 20° sidegaze, 10° and 20° downgaze, The accommodative responses was measured by a open-field autorefractometer under the distances of 3 m, 1 m and 40 cm that enable consecutive measurement. Results : In all gaze directions as distance became shorter the accommodative response increased. At each viewing distance 1m and 40 cm, the accommodative responses showed the peaks 1.09± 0.10 D in 10° and 2.04±0.11 D in 20° downgaze. respectively, while the lowest accommodative responses were 0.44±0.05 D and 1.54±0.06 D in 20° and 10° sidegaze, which showed statistical differences(p＜0.050). There were significant differences in each viewing direction except for in 10° and 20° sidegaze(p＜0.050). The gaze direction and the accommodative responses were significantly different(p＜0.050). Conclusion : This study demonstrates that the accommodative responses vary with the gaze direction and gaze angle for the same modulating stimulus. Therefore, it is assumed that 20° downgaze makes the object sharper and minimizes the eye fatigue.

Asymmetry in Gaze-Holding Impairment in Acute Unilateral Ischemic Cerebellar Lesions Critically Depends on the Involvement of the Caudal Vermis and the Dentate Nucleus.

Stabilizing the eyes in space when looking at a target is provided by a brainstem/cerebellar gaze-holding network, including the flocculus/paraflocculus complex (non-human primate studies) and the caudal vermis, biventer, and inferior semilunar lobule (human studies). Previous research suggests that acute lateralized cerebellar lesions preferentially lead to gaze-evoked nystagmus (GEN) on ipsilesional gaze. Here, we further characterize the effect of unilateral cerebellar lesions on gaze-holding and hypothesize that the side-specific magnitude of gaze-holding impairment depends on the lesion location. Nine patients (age range = 31-62 years) with acute/subacute (≤ 10 days old) MRI-confirmed unilateral cerebellar stroke were included. Horizontal gaze holding was quantified while looking at a slowly moving (0.5°/s) flashing target (gaze angle = ±40°). Asymmetry in eye-drift velocity was calculated and compared with the different MRI patterns of cerebellar lesions. Individual peak eye-drift velocities (range = 1.7-8.8°/s) occurred at the most eccentric eye positions (gaze angle = 28-38°). We found significantly asymmetric eye-drift velocity (EDV) in eight out of nine patients. The four patients with MRI-confirmed involvement of the caudal vermis and the dentate nucleus all presented with ipsilesionally-predominant EDV, while in the five patients with lesions restricted to the cerebellar hemisphere, EDV was stronger on contralesional gaze in three out of four found with an asymmetric EDV. Involvement of the caudal vermis and the dentate nucleus is critical for determining the directional GEN asymmetry in unilateral cerebellar lesions. Thus, our findings support the occurrence of GEN without floccular/parafloccular lesions and suggest that the EDV asymmetry in relation to the side of the lesion provides information about the involvement of specific structures.

Expanding the Bounds of Seated Virtual Workspaces

Mixed Reality (MR), Augmented Reality (AR) and Virtual Reality (VR) headsets can improve upon existing physical multi-display environments by rendering large, ergonomic virtual display spaces whenever and wherever they are needed. However, given the physical and ergonomic limitations of neck movement, users may need assistance to view these display spaces comfortably. Through two studies, we developed new ways of minimising the physical effort and discomfort of viewing such display spaces. We first explored how the mapping between gaze angle and display position could be manipulated, helping users view wider display spaces than currently possible within an acceptable and comfortable range of neck movement. We then compared our implicit control of display position based on head orientation against explicit user control, finding significant benefits in terms of user preference, workload and comfort for implicit control. Our novel techniques create new opportunities for productive work by leveraging MR headsets to create interactive wide virtual workspaces with improved comfort and usability. These workspaces are flexible and can be used on-the-go, e.g., to improve remote working or make better use of commuter journeys.

0030 Effect of a Simulated Sunset versus Typical Indoor Lighting on Evening Melatonin Levels

Abstract Introduction The human circadian clock evolved in the presence of the natural light-dark solar cycle. Exposure to artificial light at night suppresses endogenous melatonin levels and delays the timing of the circadian clock. The advent of tunable LED (light emitting diode) technology presents an opportunity to develop and implement circadian based practices for healthy lighting. Here we determined the influence of a simulated sunset with tunable LED technology on evening melatonin levels. Methods Nine healthy adults (3 females, 24.0 ± 5.3 years) completed a 15-day crossover study comparing typical artificial indoor lighting versus a simulated sunset using tunable LEDs (Acuity Brands-Rubik). After 1 week maintaining an ~8h sleep schedule, participants arrived at the laboratory 6h prior to habitual bedtime. Typical electrical indoor home lighting was &amp;lt;90 lux in angle of gaze until scheduled bedtime (&amp;lt;467 lux maximum at 183 cm in the direction of the ceiling mounted light fixtures; 3,500K). To simulate sunset, a simple least square fit was used to match relative spectral irradiance of the laboratory LED lighting to solar spectral irradiance of a standard mid-latitude summer atmosphere in Boulder, Colorado with solar elevation angles ranging from 3.9 degrees to 0 degrees (sunset). The first 3h30min of the simulation was typical indoor lighting of &amp;lt;90 lux (angle of gaze; 3,500K) followed by a 25 min transition in spectral irradiance and then 2h5min at ~7 lux in the angle of gaze (&amp;lt;38 lux maximum at 183 cm; 2,700K). Results Melatonin levels were initially similar between conditions but were significantly higher (p&amp;lt;0.05) after the sunset transition in the simulated sunset condition compared to the typical electrical indoor home lighting condition. Conclusion These preliminary findings suggest that simulating a sunset transition with tunable LED technology prior to habitual bedtime in the evening has potential to benefit circadian health. Support This work was supported in part by NIH R01 HL135598 and NASA Award 80NSSC17K0569.

Accurate Real‐time 3D Gaze Tracking Using a Lightweight Eyeball Calibration

Abstract3D gaze tracking from a single RGB camera is very challenging due to the lack of information in determining the accurate gaze target from a monocular RGB sequence. The eyes tend to occupy only a small portion of the video, and even small errors in estimated eye orientations can lead to very large errors in the triangulated gaze target. We overcome these difficulties with a novel lightweight eyeball calibration scheme that determines the user‐specific visual axis, eyeball size and position in the head. Unlike the previous calibration techniques, we do not need the ground truth positions of the gaze points. In the online stage, gaze is tracked by a new gaze fitting algorithm, and refined by a 3D gaze regression method to correct for bias errors. Our regression is pre‐trained on several individuals and works well for novel users. After the lightweight one‐time user calibration, our method operates in real time. Experiments show that our technique achieves state‐of‐the‐art accuracy in gaze angle estimation, and we demonstrate applications of 3D gaze target tracking and gaze retargeting to an animated 3D character.