Vision and Gaze: A Guide to Eye Tracking Technology

dbacic47
Apr 12
4 min read

Author: Benjamin Liddell

Eye tracking has become a valuable biometric tool for measuring engagement with types of stimuli. Eye tracking technology is a type of sensor that tracks the position and movement of the eyes. All types of eye trackers have the same main functionality: they measure gaze points and fixation duration to assess attention or engagement. While all eye trackers have the same purpose, they don’t all work the same way. The way they are designed determines where they can be used and the quality of the data they provide. There are two main categories of eye trackers: remote or screen-based trackers and wearable or mobile-based trackers. In each category, there are different types of eye trackers

Remote or Screen-Based Eye Tracker:

Remote-based eye trackers are the most popular because they are more widely available than mobile-based ones. These are called remote because the user does not touch or wear the device; they only look at it. These eye trackers work by mounting a camera and a light sensor on either the top or bottom of the device to create the headbox. The headbox is a virtual space where the user’s eyes can be tracked while allowing that person’s head to move naturally. These spaces can be big or small, depending on the type of tracker used. These headboxes are important for remote-based eye trackers because, without them, users wouldn’t be able to move their heads at all during the session.

Standalone eye trackers

Within the remote-based eye tracker category, there are different types of trackers that can be used depending on business needs. The ﬁrst one’s are standalone eye trackers. These are more common in advanced research. Standalone eye trackers use the pupil-center corneal-reflection method. This method works by having a camera look at the pupil while a light source, usually an infrared (IR) sensor, illuminates the eye. That light creates a glint, which is the reﬂection on the cornea. The software then uses that data to calculate the distance between the pupil and the glint. The eye tracker can then use that data to calculate gaze points on the stimuli. While physical eye trackers provide highly accurate data, they are not always practical for large-scale studies. In large-scale studies where it is not feasible to have everyone come to a lab, webcam-based trackers excel.

Since it requires no additional hardware beyond a webcam and internet access, which most devices have, this type of eye tracker is the most accessible. They work differently from physical remote eye trackers; instead of using the pupil-center corneal-reflection method, they primarily rely on computer vision. While standalone trackers have dedicated light sensors to track movement, webcam trackers use the light sensors built into the webcam. Since they rely on the amount of lighting in the room, their accuracy depends on the lighting conditions. They won’t be as reliable if there is too much or too little lighting. Another difference is the refresh rates webcam eye trackers use. They can only use the refresh rates built into the device, which are about 30-60 Hz, whereas standalone eye trackers often start at 60 Hz and can go much higher than that. Using webcams is still as important to researchers because, even though they are not as precise, they are better at doing studies that require a lot of participants.

Integrated eye trackers

A newer type of eye tracker is called an integrated eye tracker, and it is more commonly found in vehicles for active driving systems, such as General Motors’ Super Cruise and Ford Motor Company’s Blue Cruise. These eye trackers are a combination of both webcam and standalone trackers. They use the technology found in standalone trackers to collect data for decision-making, but leverage the hardware and scalability of webcams. These trackers are always mounted on the steering wheel or dashboard, so they are always looking at the driver. It uses infrared (IR) sensors, employing a method like the pupil center corneal reflection method, to track eye gaze in all sorts of weather conditions. Once the system detects the eyes wandering from the road, the car starts playing warning sounds until corrective action is taken. This approach allows the system to continuously track the driver, ensure they are paying attention to the road, and take corrective action if needed.

Wearable or Mobile-Based Eye Trackers

Wearable eye trackers have become more popular because they allow researchers to conduct eye tracking in real-world settings rather than in a lab. There are two types of trackers in use: Eyeglasses and Virtual Reality Goggles. These types of trackers have different purposes: one is for real-world research, and the other is for virtual-world research.

Eyeglasses

The eyeglasses are worn exactly like regular eyeglasses; they have cameras on the outside, usually on the bridge, to provide a first-person point of view. Inside the glasses, there are infrared (IR) sensors that track pupil movement. Using data from the IR sensor and the outward-facing camera, the computer can generate gaze points to indicate what the person is looking at. This type of wearable is used primarily for marketing studies that require how a person may look at something in the real world.

Virtual Reality Goggles

Another type of wearable eye tracker is virtual reality goggles. Unlike the eyeglass tracker, which requires people to use it in the real world, the virtual reality goggles are used in a virtual world. The Apple Vision Pro and the Meta Quest use virtual reality to function. They have a row of IR sensors and eye cameras that constantly track eye movements.

That data is sent to the onboard computer, which can create gaze points to pinpoint where the person is looking. When combined with hand tracking using external cameras, users can control the device with gaze and gestures. The Apple Vision Pro differs from other virtual reality devices because it uses data to recognize each person’s unique eye, allowing them to unlock the device. These eye trackers are helpful when researchers need to control a real-life scenario or when people need to do job-related simulation training.

Conclusions

Eye trackers are becoming increasingly important with the rise in popularity of biometrics. The type of eye tracker chosen matters for each type of study and environment. If the study is about how someone looks at things in the real world, mobile eye trackers should be used because it’s hard to replicate real-world conditions with a remote tracker. If the study is about how users interact with stimuli in a controlled, computer-based setting, remote eye trackers should be used because they provide high-quality data. By understanding the study's goal and the required data accuracy, researchers can choose the right eye tracker.