| Abstract |
| This study aims to develop a novel interface technology based on the pupillary light reflex (PLR) using visual stimuli incorporating dynamic brightness changes. Ten types of visual stimuli featuring variations in luminance between 0 and 255 were designed. Pupil size changes were measured and analyzed through a task in which subjects focused on the visual stimuli for 3 seconds. One trial was defined as the presentation of all ten visual stimuli in a random order. The experiment consisted of 12 trials, with each visual stimulus presented 120 times in total. Among five deep learning-based sequence models, the temporal convolutional network exhibited the highest performance across ten subjects, achieving a classification accuracy of 94.01 ± 3.94% and an information transfer rate of 56.64 ± 6.01 bits/min. The PLR-based interface is an intuitive system that does not require user training. Its high scalability, facilitated by the development of diverse visual stimulus patterns, makes it a promising technology for future applications. |
|
|
| Key Words |
| 동공 반응, 동공 빛 반사, 인간-컴퓨터 상호작용, 사용자 인터페이스, 시각 자극, Pupillary Response, Pupillary Light Reflex (PLR), Human-Computer Interaction (HCI), User Interface (UI), Visual Stimulus |
|
|
 |
|
