Head-Mounted Eye Tracking of a Chimpanzee under Naturalistic Conditions
Fumihiro Kano, Masaki Tomonaga
The eye tracker on Pan’s head (A) and the eye (left top) and scene-camera image (B). Cross mark indicates point-of-regard (POR).
Scene-camera image with point-of-regard (POR; cross mark). The first half of image was recorded during the greeting phase, and the latter half was recorded during the task phase.
This study offers a new method for examining the bodily, manual, and eye movements of a chimpanzee at the micro-level. A female chimpanzee wore a lightweight head-mounted eye tracker (60 Hz) on her head while engaging in daily interactions with the human experimenter. The eye tracker recorded her eye movements accurately while the chimpanzee freely moved her head, hands, and body. Three video cameras recorded the bodily and manual movements of the chimpanzee from multiple angles. We examined how the chimpanzee viewed the experimenter in this interactive setting and how the eye movements were related to the ongoing interactive contexts and actions. We prepared two experimentally defined contexts in each session: a face-to-face greeting phase upon the appearance of the experimenter in the experimental room, and a subsequent face-to-face task phase that included manual gestures and fruit rewards. Overall, the general viewing pattern of the chimpanzee, measured in terms of duration of individual fixations, length of individual saccades, and total viewing duration of the experimenter’s face/body, was very similar to that observed in previous eye-tracking studies that used non-interactive situations, despite the differences in the experimental settings. However, the chimpanzee viewed the experimenter and the scene objects differently depending on the ongoing context and actions. The chimpanzee viewed the experimenter’s face and body during the greeting phase, but viewed the experimenter’s face and hands as well as the fruit reward during the task phase. These differences can be explained by the differential bodily/manual actions produced by the chimpanzee and the experimenter during each experimental phase (i.e., greeting gestures, task cueing). Additionally, the chimpanzee’s viewing pattern varied depending on the identity of the experimenter (i.e., the chimpanzee’s prior experience with the experimenter). These methods and results offer new possibilities for examining the natural gaze behavior of chimpanzees.