Eye Gaze Accuracy Research Articles

Head Movement Invariant Eye Tracking System

The human visual framework, consisting of the eyes, optic nerve, and brain, gives us the sense of sight to capture and interact with our surroundings. The eye, which provides stimulus to generate sight, is a complex organ that collects light and effectively converts it to an electro-chemical impulse inside neurons.Eye trackers are an important input modality for human-computer interfaces with vision-based applications. Eye-tracking systems have been proven to be valuable assistive devices for people with motor disabilities. Despite significant improvements in eye-tracking technology over time, current eye trackers in the market are complex and require intricate calibration procedures. Hence, there is still room for further research to develop an eye-tracking system with simpler calibration procedures.This paper presents the development of a head movement invariant real-time eye-tracking system with simpler calibration procedures that can operate non-intrusively. The complexity of setting up the system is also significantly reduced. Experiments conducted with a sample set of users showed an eye gaze accuracy improvement of over 11% compared to the existing setups commonly available in the market.

Effect of feedback and target size on eye gaze accuracy in an off-screen task

Purpose Eye gaze interfaces have been used by people with severe physical impairment to interact with various assistive technologies. If used to control robots, it would be beneficial if individuals could gaze directly at targets in the physical environment rather than have to switch their gaze between a screen with representations of robot commands and the physical environment to see the response of their selection. By using a homogeneous transformation technique, eye gaze coordinates can be mapped between the reference coordinate frame of eye tracker and the coordinate frame of objects in the physical environment. Feedback about where the eye tracker has determined the eye gaze is fixated is needed so users can select targets more accurately. Screen-based assistive technologies can use visual feedback, but in a physical environment, other forms of feedback need to be examined. Materials and methods In this study, an eye gaze system with different feedback conditions (i.e., visual, auditory, vibrotactile, and no-feedback) was tested when participants received visual feedback on a display (on-screen) and when looking directly at the physical environment (off-screen). Target selection tasks in both screen conditions were performed by ten non-disabled adults, three non-disabled children, and two adults and one child with cerebral palsy. Results Tasks performed with gaze fixation feedback modalities were accomplished faster and with higher success than tasks performed without feedback, and similar results were observed in both screen conditions. No significant difference was observed in performance across the feedback modalities, but participants had personal preferences. Conclusion The homogeneous transformation technique enabled the use of a stationary eye tracker to select target objects in the physical environment, and auditory and vibrotactile feedback enabled participants to be more accurate selecting targets than without it. Implications for Rehabilitation Being able to select target objects in the physical environment by eye gaze could make it easier for children with disabilities to control assistive robots, because in this way they do not have to change their focus between a computer screen with commands and the robot. Providing auditory or vibrotactile feedback when using an eye gaze system made it faster and easier to know if a target was being gazed upon. Being able to select targets in the environment using eye gaze could be beneficial for other assistive technology, too, such as destination selection for power wheelchairs.

Eye movement measures for predicting eye gaze accuracy and symptoms in 2D and 3D displays

The current study applied Structural Equation Modeling (SEM) to analyze the interrelationship among index of difficulty (ID), environment, saccade duration (SD), revisited fixation duration (RFD), number of fixation (NF), pupil size (PS), eye gaze accuracy (AC), and symptoms simultaneously. SD, RFD, NF, PS, and AC were measured by utilizing the Tobii eye tracker system. Twelve participants were recruited to perform multi-directional tapping task using within-subject design with two different environments (2D screen display and 3D stereoscopic display) and six different levels of ID. SEM showed that ID had significant direct effects on SD and RFD while environment was found had significant direct effects on SD, RFD, PS, AC, and symptoms. Among selected eye movement measures, NF was found to be the best predictor of AC and PS was found to be the best predictor of symptoms. In addition, RFD was also found to be a good predictor of symptoms. Our results found that higher AC was achieved by projecting the image in the 2D screen display with higher ID and it resulted in higher SD and higher NF. Regarding the symptoms, our results found that lower symptoms were achieved by projecting the image in the 2D screen display with lower ID and it resulted in lower PS, and lower RFD.Practitioner Summary: The SEM could provide valuable theoretical foundations to identify the interrelationship among eye movement measures, AC, and symptoms particularly for VR researchers and interface developers.

Convolutional Neural Network Implementation for Eye-Gaze Estimation on Low-Quality Consumer Imaging Systems

Accurate and efficient eye gaze estimation is important for emerging consumer electronic systems, such as driver monitoring systems and novel user interfaces. Such systems are required to operate reliably in difficult, unconstrained environments with low power consumption and at minimal cost. In this paper, a new hardware friendly, convolutional neural network (CNN) model with minimal computational requirements is introduced and assessed for efficient appearance-based gaze estimation. The model is tested and compared against existing appearance-based CNN approaches, achieving better eye gaze accuracy with significantly fewer computational requirements.

Eye Movement Parameters for Performance Evaluation in Projection-based Stereoscopic Display.

The current study applied Structural Equation Modeling (SEM) to analyze the rela-tionship among index of difficulty (ID) and parallax on eye gaze movement time (EMT), fixation duration (FD), time to first fixation (TFF), number of fixation (NF), and eye gaze accuracy (AC) simultaneously. EMT, FD, TFF, NF, and AC were measured in the projec-tion-based stereoscopic display by utilizing Tobii eye tracker system. Ten participants were recruited to perform multi-directional tapping task using within-subject design with three different levels of parallax and six different levels of ID. SEM proved that ID had significant direct effects on EMT, NF, and FD also a significant indirect effect on NF. However, ID was found not a strong predictor for AC. SEM also proved that parallax had significant direct effects on EMT, NF, FD, TFF, and AC. Apart from the direct effect, parallax also had significant indirect effects on NF and AC. Regarding the interrelation-ship among dependent variables, there were significant indirect effects of FD and TFF on AC. Our results concluded that higher AC was achieved by lowering parallax (at the screen), longer EMT, higher NF, longer FD, and longer TFF.Practitioner Summary: The SEM could provide valuable theoretical foundations of the interrelationship among eye movement parameters for VR researchers and human-virtual-reality interface developers especially for predicting eye gaze accuracy.

Using computer‐assisted instruction to increase the eye gaze of children with autism

Many children diagnosed with autism spectrum disorders have difficulty making appropriate eye contact and engaging in joint attention. The current study evaluated a computer‐assisted instruction package (pairing visual stimuli with vocal stimuli) as a novel treatment to improve the eye gaze accuracy in 3 elementary school children with autism. The researchers measured the latency from a recorded verbal stimulus to the students making eye contact with pictures of familiar individuals displayed on a computer screen, and the duration for which eye gaze on the stimulus was maintained. An automated infrared camera system for measuring eye gaze was utilized that eliminated the need for an instructor to make subjective judgments regarding participants' eye gaze. For all three participants, duration of eye contact increased, and latency to responding decreased following exposure to the computer‐assisted instruction. The implications of these findings for the treatment of individuals with autism are discussed, along with suggestions for future research on the topic.