Abstract
Intro: Typical development of basic visual functions depends on active self-motion through a visual environment (as opposed to passive viewing), yet whether self-motion impacts visual scene perception in adults is unknown. To find out, we tested memory for real-world panoramic scenes studied in head-mounted virtual reality with either with active self-motion or passive viewing. Methods: 24 participants viewed 36 novel, real-world 360º photospheres through a headmounted Virtual Reality Display (Oculus Rift). Scenes were presented through the Oculus Rift in three conditions, in which participants: (1) actively explored through the use of head turns (Oculus Condition), (2) actively explored using a joystick to change viewpoints while head position remained fixed (Joystick Condition), or (3) passively viewed "Playback" recordings of active exploration trials by different individuals (Video Condition). Subsequently, three memory tests were administered. Memory Test 1, Recognition Memory: Participants were asked to recall whether they had viewed a snapshot showing a 90º subset of a studied photosphere. Memory Test 2, Panoramic Memory: Participants judged the relative spatial positioning of two scene snapshots. Memory Test 3, Spatial Memory: Participants judged whether a scene snapshot had appeared on the left or right side of the scene. Results: We observed a main effect of study condition on memory performance (F(2,44)=13.6, p=0.01): overall, self-directed, first-person experience increased subsequent scene memory over passive viewing (Oculus > Joystick > Video Condition). We also observed an interaction between study condition and memory test (F(4,88)=9.5, p=0.01): Performance on Memory Test 2 was particularly enhanced by fully active viewing (Oculus > Joystick, p=0.038). Conclusions: Our results demonstrate that active engagement with a panoramic environment parametrically benefits subsequent scene memory. The strength of this benefit for panoramic memory in particular suggests that panoramic memory formation may be facilitated by a mapping between viewpoints and head-orientation.
Meeting abstract presented at VSS 2017