The role of vision in locomotion
dc.contributor.advisor | Hayhoe, Mary | |
dc.contributor.committeeMember | Cormack, Lawrence K | |
dc.contributor.committeeMember | Yu, Chen | |
dc.contributor.committeeMember | Bonnen, Kathryn | |
dc.creator | Panfili, Daniel P. | |
dc.creator.orcid | 0009-0000-0922-5095 | |
dc.date.accessioned | 2024-01-03T23:15:15Z | |
dc.date.available | 2024-01-03T23:15:15Z | |
dc.date.created | 2023-08 | |
dc.date.issued | 2023-08-11 | |
dc.date.submitted | August 2023 | |
dc.date.updated | 2024-01-03T23:15:16Z | |
dc.description.abstract | The human visual system depends on the ability to move the eyes in order to provide a high-acuity signal for evaluating and exploring the environment. However, much of the existing research on vision and locomotion has been performed in highly constrained settings, presenting 2-dimensional images on a screen to head-fixed or treadmill-bound walkers. Head-bob and active stabilization of the visual image are characteristic of bipedal locomotion, and these elements of often absent or estimated in standard experimental paradigms. Because the retinal image changes as the eye and body move in the environment, and because the new image provides information for the next action, it is necessary to measure eye and body movements together with visual scene information in the context of natural behavior in order to understand the problems that the visual system needs to solve. In this dissertation, we used recently developed photogrammetry algorithms on data collected from walks in outdoor terrains to reconstruct a 3D scene representation. This allowed for detailed analysis of the visual control of locomotion, specifically how gaze is directed when searching for footholds in complex terrain. The terrain representations allow us to localize the body and eye in the environment, giving much more accurate analyses on the relationship between gaze and footholds. The data suggests that the terrain up to 5 steps ahead is used to plan future step locations, and gaze distribution depends on individual subject kinematics. This new form of analysis also allowed inferences about different potential roles for gaze in locomotor control. Following on from previous work (Matthis et al. 2022), we explored the role of optic flow in controlling the direction of travel. Previous work (Rushton et al. 1999, Warren et al. 2001) has focused on the focus of expansion (FOE) as the primary visual cue in controlling direction. Our experiments in a virtual environment suggest that retinal flow patterns influence directional control locally, and visual direction cues are likely used to steer toward distant targets. Additionally, motion in the lower visual field appears to be more influential than the upper visual field in local path selection. An extension of this paradigm was used to explore the influence of gaze direction, showing that paths are robust to choice of gaze location. We show that the retinal motion patterns appear to be influencing the paths taken, and previous work has shown that flow patterns have a strong influence of posture on stepping. To verify whether individuals use this information during locomotion to maintain balance and control stepping, we ask if subjects learn the flow patterns consistent with the passive dynamics that determine their preferred gait. We manipulated the retinal flow patterns generated during locomotion, and were able to directly influence the gait parameters. This suggests that subjects learn their self-generated flow patterns and use this as a postural control signal. The manipulations also affected gaze allocation, revealing the connection between gaze behavior and foot placement, regulated by visual uncertainty. These experiments reveal the importance of understanding the visual stimulus generated during active behavior in the natural world. | |
dc.description.department | Psychology | |
dc.format.mimetype | application/pdf | |
dc.identifier.uri | https://hdl.handle.net/2152/123321 | |
dc.identifier.uri | https://doi.org/10.26153/tsw/50119 | |
dc.language.iso | en | |
dc.subject | Vision | |
dc.subject | Locomotion | |
dc.subject | Photogrammetry | |
dc.subject | Virtual reality | |
dc.subject | Optic flow | |
dc.title | The role of vision in locomotion | |
dc.type | Thesis | |
dc.type.material | text | |
thesis.degree.department | Psychology | |
thesis.degree.discipline | Psychology | |
thesis.degree.grantor | The University of Texas at Austin | |
thesis.degree.level | Doctoral | |
thesis.degree.name | Doctor of Philosophy |
Access full-text files
Original bundle
1 - 1 of 1
Loading...
- Name:
- PANFILI-DISSERTATION-2023.pdf
- Size:
- 7.83 MB
- Format:
- Adobe Portable Document Format