Eccentric perception of biological motion is unscalably poor

Abstract

Accurately perceiving the activities of other people is a crucially important social skill of obvious survival value. Human vision is equipped with highly sensitive mechanisms for recognizing activities performed by others [Johansson, G. (1973). Visual perception of biological motion and a model for its analysis. Perception and Psychophysics, 14, 201; Johansson, G. (1976). Spatio-temporal differentiation and integration in visual motion perception: An experimental and theoretical analysis of calculus-like functions in visual data processing. Psychological Research, 38, 379]. One putative functional role of biological motion perception is to register the presence of biological events anywhere within the visual field, not just within central vision. To assess the salience of biological motion throughout the visual field, we compared the detectability performances of biological motion animations imaged in central vision and in peripheral vision. To compensate for the poorer spatial resolution within the periphery, we spatially magnified the motion tokens defining biological motion. Normal and scrambled biological motion sequences were embedded in motion noise and presented in two successively viewed intervals on each trial (2AFC). Subjects indicated which of the two intervals contained normal biological motion. A staircase procedure varied the number of noise dots to produce a criterion level of discrimination performance. For both foveal and peripheral viewing, performance increased but saturated with stimulus size. Foveal and peripheral performance could not be equated by any magnitude of size scaling. Moreover, the inversion effect--superiority of upright over inverted biological motion [Sumi, S. (1984). Upside-down presentation of the Johansson moving light-spot pattern. Perception, 13, 283]--was found only when animations were viewed within the central visual field. Evidently the neural resource responsible for biological motion perception are embodied within neural mechanisms focused on central vision.

Perception of Biological Motion in Central and Peripheral Visual Fields

“…Object size and chosen eccentricities were similar to parameters used in Ikeda et al (2005) and Gurnsey et al (2008). Point size was scaled proportionally to the object size.…”

“…When demonstrating a pointlight-walker embedded in motion noise, consisting of dots with the same physical properties (including dot size, motion trajectories, etc.) as the dots forming the biological object and thus preserving the local motion information, the observer is still capable of distinguishing between the object of biological motion and the noise, by using the information about the whole structure of the object (Cutting et al, 1988;Bertenhal and Pinto, 1994;Giese and Lappe, 2002;Ikeda et al, 2005;Freire et al, 2006;Schouten et al, 2013).…”

“…Analogously to the previously mentioned studies that analysed visual acuity, face perception, etc., studies analysing the sensitivity to biological motion also use stimulus magnification (Ikeda et al, 2005;Gurnsey et al, 2008;2010). However, the results of these studies are quite ambiguous.…”

“…However, the results of these studies are quite ambiguous. Ikeda et al (2005) described the eccentric perception of biological motion as "unscalably poor", showing that when the stimulus ass embedded in motion noise the discrimination of a point light walker from its scrambled version was more effective in the central visual field and stimulus magnification did not compensate for reduced performance in the peripheral visual field. On the other hand, Gurnsey et al (2008) come to completely opposite results by analysing the facing direction of a biological object and demonstrated that increase in stimulus size equalised the performance in both central and peripheral visual fields.…”

“…Although both of these studies used a similar experimental setup, their results and main conclusions seem to be entirely opposite. Gurnsey et al (2008) argued that the differences might have been related to the fact that discriminating an object from motion noise (used by Ikeda et al, 2005) requires global perception of the whole body, and determining facing direction (without additional motion noise) is more related to local analysis of motion trajectories of the feet. Hence, the functional differences between these results were not due to differences in perception of biological motion but rather due to the reduced performance in element grouping or figureground discrimination in the peripheral visual field.…”