屋内照明の角膜反射像とIMUを用いた位置推定

抄録

Corneal imaging has been studied to estimate the focused object and point-of-gaze by extracting the environmental information reflected from the corneal surface. However, it is also necessary to estimate the self-position as a daily-use device. In recent years, research has been conducted on self-position estimation using interior illumination patterns reflected on the cornea. However, localization based on illumination information was not accurate, and thus, an improvement is required. In this paper, we propose to improve the accuracy of localization by integrating SE-ResNet and IMU on the framework of Monte Carlo Localization.