Skip to content
naulon
Not checkedNot checked for notices yet.

P3-15: Perceived 3D Shape from Motion for Small and Large Perspective Changes

Young Lim Lee, Jeffrey A Saunders

i-Perception · 2012

doi:10.1068/if725 (opens in a new tab)

We email you when a new notice or a question about this paper appears.

Abstract

Studies of structure-from-motion have generally found that perceived depth from motion is not veridical and depends on speed. However, some recent studies have found that metric shape perception can be yielded by large perspective changes (Bingham and Lind, 2008 Perception & Psychophysics 70 524–540; Lee et al., 2008 Journal of Vision 8 (6) 759a). We investigated the accuracy of metric shape judgments for varied amounts of perspective change, and whether amount of perspective change modulates the speed dependence of structure-from-motion (SFM). SFM stimuli were views of rotating elliptical cylinders with varied depth-to-width ratios (0.5–1.5) and with different slants in depth (top face: 60°, 70°, or 80°). Objects were rotated around the normal axis by ±5°, 10°, 20°, or 45°. The rotation was sinusoidal with peak speed of 15 °/s, 20 °/s, or 25 °/s. We also tested a static-views control condition in which the same range of views was presented without continuous motion. We found that accuracy in the SFM conditions systemically increased with amount of change in perspective. With small perspective change, judgments showed overall underestimation of depth. Larger perspective change reduced these biases, and performance approached veridical. Judgments from static-views were much less accurate, indicating that the benefit from large perspective change is not due solely to the most extreme views. We found no effect of rotation speed even with small perspective change, contrary to expectations. Our results demonstrate that increasing perspective change can improve the accuracy of perceived 3D shape from motion, and that observers are capable of near-veridical metric judgments.

Abstract licence: CC BY

Reads

Read by people on this page

None yet.

Checked or read by AI through naulon

None yet.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.

Are you an author?

Is this your paper?

Claim it to add your note and be paid when agents cite it. Free.