The Precision of Webcam-Based Eye Tracking Techniques
In the realm of research and user experience testing, a significant shift is underway. The advent of webcam-based eye tracking, spearheaded by the latest platform, WebET 3.0, is making eye tracking more accessible and affordable than ever.
WebET 3.0, integrated into cloud-based applications like Remote Data Collection and the platform Online, allows for remote eye tracking using just a standard computer, webcam, and a stable internet connection. This makes it an ideal choice for scaling research, particularly for large-scale UX testing, A/B testing, or image/video studies with eye tracking.
The study setup for measuring the accuracy of webcam eye tracking involves presenting stimuli on a 22" computer screen in a dimly lit room, with respondents sitting in front of a neutral grey wall and at a distance of 65 cm from the web camera. Under ideal conditions, webET 3.0 demonstrates impressive results, with an average accuracy offset of 2.2 dva and only 1% of trials lost due to data dropout.
However, it's important to note that while webcam-based eye tracking offers a viable data collection tool, its accuracy is impacted by factors such as head movement, lighting, and wearing glasses compared to traditional screen-based eye trackers. For instance, data recorded from respondents who were moving and talking had an average offset of 5.0 dva, with 38% of trials having an offset larger than 5 dva.
Lower camera resolutions can also cause an increase in data offsets, with a third of trials showing an average offset above 5 dva. To optimally employ webcam eye tracking in research, respondents must adhere to best practices from the study organizer to ensure the best quality of data.
Despite these limitations, the popularity of webcam-based eye tracking stems from its ease of use, affordability, and versatility. It offers insights into human behavior and cognitive processes, and its technology is experiencing rapid advancements, driven by computer vision algorithms.
In comparison to traditional screen-based eye trackers, webcam-based solutions like WebET 3.0 offer reasonable gaze estimation accuracy for applications that do not require clinical-grade precision but perform less accurately than traditional dedicated eye trackers, particularly under diverse head movement or lighting scenarios.
While direct evaluation results for WebET 3.0 specifically are not yet widely available, insights from related technologies like WebGazer.js and consumer-grade trackers suggest that it offers reasonable gaze estimation accuracy for applications that do not require clinical-grade precision.
For research requiring both accuracy and precision, investing in proper hardware is still well worth it. However, for those seeking a more affordable entry point, webcam-based eye tracking, especially WebET 3.0, presents a viable and increasingly accurate option.
References:
[1] Duchowski, A. T. (2012). Eye tracking: a guide to methods and measurement. Springer Science & Business Media.
[5] Holmqvist, P., & Holmqvist, K. (2014). Eye tracking in usability and user experience research: a literature review. Behaviour & Information Technology, 33(3), 170-188.
Media analytics can benefit significantly from the integration of WebET 3.0's webcam-based eye tracking technology into cloud-based applications, as it provides remote access to eye tracking data using standard computer equipment and a stable internet connection. Additionally, data-and-cloud-computing techniques facilitate scaling research, with potential applications including large-scale UX testing, A/B testing, or image/video studies with eye tracking.
