Purchase this article with an account.
Martin Thirkettle, N.E. Scott-Samuel, C.P. Benton; Dances with Gabors: Contour integration and form in biological motion. Journal of Vision 2009;9(8):613. doi: 10.1167/9.8.613.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The heightened ability of humans to perceive the characteristic motion of humans and animals, even from points of light marking only the major joints, has led to suggestions of specialised perceptual mechanisms for ‘biological motion‘. However, recent work has suggested that form information, which persists in biological motion stimuli, plays an important, possibly dominant, role in processing. Biological motion perception does not rely on the local motions of individual dots; rather, it has been suggested that the motion relationships between dots are a counterpart to the static form information expressed by the configuration of those dots. This ‘opponent motion‘, expressed through the integration of a number of dots' local motion signals - especially those of limb pairs - has recently been suggested as critical to the analysis of biological motion. By creating a point light walker (PLW) composed of Gabor patches instead of dots at the major joints, and yoking the orientation of each Gabor to the path of its opponent motion, we manipulated the strength of the opponent motion signal. Using a detection-in-noise task and a novel form-degradation task, where the centroid of each dot's local motion was displaced independently, we found an advantage for orienting the carrier of the Gabor patches orthogonally to their opponent motion paths. However we found the same advantage for orthogonal orientation of Gabor carriers in static presentations of both tasks. We suggest that the improved contour integration resulting from this orthogonal carrier orientation explains the performance in our biological motion task. Rather than the characteristic animate motion, the structure conveyed by the configuration of the PLWs dots may provide the basis for processing of biological motion stimuli.
This PDF is available to Subscribers Only