Abstract
A central question in vision concerns how the statistics of natural environments shape perception. There is a growing body of evidence about how static image statistics shape perceptual processing, but relatively little is known about the statistics of motion. Self-motion generates optic flow with patterns and speeds that differ based on the motion of the body, trunk, head, or eye, the motion type (translation or rotation), and scene geometry. The statistics of optic flow due to self-motion may change across age due to development in motor abilities, head and body posture, and the relative frequency of passive versus active locomotion. Here, we assess head-centered optic flow observed by infants across a wide range of ages, postures, and scenes (e.g. indoors/outdoors) and compare the frequency of optic flow patterns and overall visual speeds in each setting. Methods. Infants wore head-mounted cameras while performing simple natural activities: Walking, sitting, playing, interacting with caregivers, riding in a stroller, or being carried in a front-facing baby carrier. We estimated optic flow from short (~30 sec) segments of the recorded videos. We used the estimated optic flow to estimate the relative frequency of optic flow patterns and visual image speeds. Results. Optic flow patterns differ between infants engaged in passive locomotion compared with those moving themselves. Infants in passive locomotion view translational flow patterns more often than actively moving infants. Further, infants who are not coupled to an adult in locomotion view a wider and slower distribution of flow speeds. Thus, the statistics of optic flow differ depending on active versus passive self-motion. Conclusion. The relative frequency of experienced flow patterns and speeds may change substantially across development. More generally, state-of-the art optic flow estimators permit a detailed analysis of the optic flow statistics of dynamic natural visual experience during development.
Meeting abstract presented at VSS 2014