Eye Gaze Angles Research Articles

A Driver Gaze Estimation Method Based on Deep Learning.

Car crashes are among the top ten leading causes of death; they could mainly be attributed to distracted drivers. An advanced driver-assistance technique (ADAT) is a procedure that can notify the driver about a dangerous scenario, reduce traffic crashes, and improve road safety. The main contribution of this work involved utilizing the driver’s attention to build an efficient ADAT. To obtain this “attention value”, the gaze tracking method is proposed. The gaze direction of the driver is critical toward understanding/discerning fatal distractions, pertaining to when it is obligatory to notify the driver about the risks on the road. A real-time gaze tracking system is proposed in this paper for the development of an ADAT that obtains and communicates the gaze information of the driver. The developed ADAT system detects various head poses of the driver and estimates eye gaze directions, which play important roles in assisting the driver and avoiding any unwanted circumstances. The first (and more significant) task in this research work involved the development of a benchmark image dataset consisting of head poses and horizontal and vertical direction gazes of the driver’s eyes. To detect the driver’s face accurately and efficiently, the You Only Look Once (YOLO-V4) face detector was used by modifying it with the Inception-v3 CNN model for robust feature learning and improved face detection. Finally, transfer learning in the InceptionResNet-v2 CNN model was performed, where the CNN was used as a classification model for head pose detection and eye gaze angle estimation; a regression layer to the InceptionResNet-v2 CNN was added instead of SoftMax and the classification output layer. The proposed model detects and estimates head pose directions and eye directions with higher accuracy. The average accuracy achieved by the head pose detection system was 91%; the model achieved a RMSE of 2.68 for vertical and 3.61 for horizontal eye gaze estimations.

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.

Foveated holographic near-eye 3D display.

We present a foveated rendering method to accelerate the amplitude-only computer-generated hologram (AO-CGH) calculation in a holographic near-eye 3D display. For a given target image, we compute a high-resolution foveal region and a low-resolution peripheral region with dramatically reduced pixel numbers. Our technique significantly improves the computation speed of the AO-CGH while maintaining the perceived image quality in the fovea. Moreover, to accommodate the eye gaze angle change, we develop an algorithm to laterally shift the foveal image with negligible extra computational cost. Our technique holds great promise in advancing the holographic 3D display in real-time use.

Self-denigration and the mixed messages of ‘ugly’ selfies in Instagram

Abstract This paper sets out a new, multimodal account (both visual and verbal analysis) of how self-denigration is conveyed through ‘ugly selfies’ as posted to the photo-sharing site Instagram. Drawing on 226 Instagram posts categorised by the poster themselves with the hashtag #uglyselfie, the visual analysis compared the persons and setting in the selfie, the person’s facial expression (eye brow position, mouth expression), eye gaze and camera angle. The verbal analysis included the self-denigration found in the image caption and the rapport enhancing or threatening responses found in the comments to the image posted by others. Ugly selfies are found to convey various types of self-denigration, including self-deprecation and self-mockery. The difference between self-deprecation and self-mockery are explained by the types of ‘mixed messages’ generated by the images. In self-deprecation, the mixed messages arise from a contrast between the visual image and the verbal caption. In self-mockery, the mixed messages arise from contrasts within the image itself, where visual exaggeration and incongruity lead to a jocular interpretation of the ugly selfie. Although the focus in this paper is on self-denigration and selfie-taking, the analysis of mixed messages can be used to analyse many different forms of politeness strategies and many kinds of images shared with increasing frequency and importance on social media sites.

Assessing the Impact of Corneal Refraction and Iris Tissue Non-Planarity on Iris Recognition

Assumptions regarding eye morphology are implicit in iris recognition algorithms. The cornea is assumed to have little to no effect on the view of the iris texture when the eye gaze is non-frontal, the iris is assumed flat, and the eye is assumed to be imaged via an orthographic projection. If these assumptions hold, affine transformations may be used to rectify non-frontally posed images to a frontal view and the rubber sheet model may be accurately used to normalize the iris annulus to a rectangular image. This paper examines how iris recognition performance degrades when the first two assumptions are violated. Using a computer renderable eye model, a data set is created varying the presence of the cornea and a parameterized non-planarity of the iris shape across a large range of eye gaze angles. Matching scores are created using a commercial matcher. When comparing the relative impact of each assumption violation, it is observed that iris non-planarity presents a more significant problem than corneal refractive distortion with regard to iris recognition accuracy in non-frontal images.

Continuous Driver's Gaze Zone Estimation Using RGB-D Camera.

The driver gaze zone is an indicator of a driver’s attention and plays an important role in the driver’s activity monitoring. Due to the bad initialization of point-cloud transformation, gaze zone systems using RGB-D cameras and ICP (Iterative Closet Points) algorithm do not work well under long-time head motion. In this work, a solution for a continuous driver gaze zone estimation system in real-world driving situations is proposed, combining multi-zone ICP-based head pose tracking and appearance-based gaze estimation. To initiate and update the coarse transformation of ICP, a particle filter with auxiliary sampling is employed for head state tracking, which accelerates the iterative convergence of ICP. Multiple templates for different gaze zone are applied to balance the templates revision of ICP under large head movement. For the RGB information, an appearance-based gaze estimation method with two-stage neighbor selection is utilized, which treats the gaze prediction as the combination of neighbor query (in head pose and eye image feature space) and linear regression (between eye image feature space and gaze angle space). The experimental results show that the proposed method outperforms the baseline methods on gaze estimation, and can provide a stable head pose tracking for driver behavior analysis in real-world driving scenarios.

Real-time estimation of horizontal gaze angle by saccade integration using in-ear electrooculography.

The manuscript proposes and evaluates a real-time algorithm for estimating eye gaze angle based solely on single-channel electrooculography (EOG), which can be obtained directly from the ear canal using conductive ear moulds. In contrast to conventional high-pass filtering, we used an algorithm that calculates absolute eye gaze angle via statistical analysis of detected saccades. The estimated eye positions of the new algorithm were still noisy. However, the performance in terms of Pearson product-moment correlation coefficients was significantly better than the conventional approach in some instances. The results suggest that in-ear EOG signals captured with conductive ear moulds could serve as a basis for light-weight and portable horizontal eye gaze angle estimation suitable for a broad range of applications. For instance, for hearing aids to steer the directivity of microphones in the direction of the user’s eye gaze.

Trait Anxiety Impacts the Perceived Gaze Direction of Fearful But Not Angry Faces.

Facial expression and gaze direction play an important role in social communication. Previous research has demonstrated the perception of anger is enhanced by direct gaze, whereas, it is unclear whether perception of fear is enhanced by averted gaze. In addition, previous research has shown the anxiety affects the processing of facial expression and gaze direction, but hasn’t measured or controlled for depression. As a result, firm conclusions cannot be made regarding the impact of individual differences in anxiety and depression on perceptions of face expressions and gaze direction. The current study attempted to reexamine the effect of the anxiety level on the processing of facial expressions and gaze direction by matching participants on depression scores. A reliable psychophysical index of the range of eye gaze angles judged as being directed at oneself [the cone of direct gaze (CoDG)] was used as the dependent variable in this study. Participants were stratified into high/low trait anxiety groups and asked to judge the gaze of angry, fearful, and neutral faces across a range of gaze directions. The result showed: (1) the perception of gaze direction was influenced by facial expression and this was modulated by trait anxiety. For the high trait anxiety group, the CoDG for angry expressions was wider than for fearful and neutral expressions, and no significant difference emerged between fearful and neutral expressions; For the low trait anxiety group, the CoDG for both angry and fearful expressions was wider than for neutral, and no significant difference emerged between angry and fearful expressions. (2) Trait anxiety modulated the perception of gaze direction only in the fearful condition, such that the fearful CoDG for the high trait anxiety group was narrower than the low trait anxiety group. This demonstrated that anxiety distinctly affected gaze perception in expressions that convey threat (angry, fearful), such that a high trait anxiety level modulated the impact of indirectly threatening expressions (fearful), and did not influence responses to directly threatening expression (angry). These findings partially support the shared signal hypothesis.

The effects of the methods of eye gaze and visual angles on accuracy of P300 speller

The effects of the methods of eye gaze and visual angles on accuracy of P300 speller

Cone of direct gaze as a marker of social anxiety in males

The fear of being scrutinised is a core feature of social anxiety disorder and socially anxious individuals overestimate being ‘looked at’. A recent development in the vision sciences is a reliable psychophysical index of the range of eye gaze angles judged as being directed at oneself (Cone of Direct Gaze: CoDG). We tested the CoDG as a measure of “being looked at” in social anxiety. Participants were stratified into high/low social anxiety groups and asked to judge whether they were being ‘looked at’ by computerised male faces varying in eye gaze deviation and facial emotion. High socially anxious males had a wider CoDG than low socially anxious males; high and low socially anxious females did not differ. Fearful faces elicited narrower cones than neutral or angry faces; however, the effect size was small and not evident for the high socially anxious males. Measures of subjective reactions to the study, and to being looked at in general, indicated that the results may be in part due to males suffering more discomfort when being looked at. The results show that measures derived from psychophysics, in this case, the CoDG, have potential as clinical and research tools for measuring anxiety about being scrutinised.

Eyegaze Detection from Monocular Camera Image for Eyegaze Communication System

An eyegaze interface is one of the key technologies as an input device in the ubiquitous-computing society. In particular, an eyegaze communication system is very important and useful for severely handicapped users such as quadriplegic patients. Most of the conventional eyegaze tracking algorithms require specific light sources, equipment and devices. In this study, a simple eyegaze detection algorithm is proposed using a single monocular video camera. The proposed algorithm works under the condition of fixed head pose, but slight movement of the face is accepted. In our system, we assume that all users have the same eyeball size based on physiological eyeball models. However, we succeed to calibrate the physiologic movement of the eyeball center depending on the gazing direction by approximating it as a change in the eyeball radius. In the gaze detection stage, the iris is extracted from a captured face frame by using the Hough transform. Then, the eyegaze angle is derived by calculating the Euclidean distance of the iris centers between the extracted frame and a reference frame captured in the calibration process. We apply our system to an eyegaze communication interface, and verified the performance through key typing experiments with a visual keyboard on display.

Perception of eye contact in video teleconsultation

During patient consultations by videoconferencing, clinicians often sit as close as 1 m from the videoconferencing units, creating a significant eye gaze angle (i.e. the angle between the eye and the camera, and the eye and the centre of the display). Eye gaze angle may adversely affect the satisfaction with videoconferencing. Four videoconferencing environments were examined: desktop PC, portable telehealth unit, videoconferencing room, and a boardroom equipped with a ceiling-mounted camera and a projection screen. Two still images of each of the three subjects were captured: one at a 7 degrees eye gaze angle and the other at 15 degrees. Each of 53 observers ranked four pairs of images for each of the three subjects. In 87% of cases, the observers perceived better eye contact at an eye gaze angle of 7 degrees than 15 degrees. Also, 92% of observers responded that the difference in the perceived eye contact was important to them as patients. Improved eye contact can be realized by increasing the horizontal distance of participants from the videoconferencing unit.