Abstract
Purpose: The ultimate aim of this study is to elucidate the mechanism inducing perceptual filling-in from the viewpoint of spatio-temporal frequency characteristics in vision. Time to filling-in is not a suitable criterion for representing facilitation of filling-in because filling-in time does not distinguish among cases: filling-in is rapidly induced by a certain mechanism and a filling-in target is essentially difficult to distinguish from its surrounding. Therefore, we introduced a new criterion - the attenuation factor of distinguishability of a filling-in target from its surroundings - to represent facilitation of filling-in (Yokota, ARVO 2004). This study investigates the attenuation factor as a function of the spatio-temporal frequency of surrounding dynamic textures. Method: Filling-in time was measured for various dynamic textures characterized by different spatio-temporal frequencies (EX1). An annular filling-in target centered at a fixation point was adopted to stabilize the subject's fixation. The target was presented at 17 [deg] eccentricity of the subject's monocular visual field. We measured visual sensitivity for each dynamic texture used in EX1 (EX2). Results: The attenuation factor, which reflects facilitation of filling-in, was estimated using the measured filling-in time and visual sensitivity for each dynamic texture. The estimated attenuation factor, when represented as a function of spatio-temporal frequency, has a bimodal shape. Those peak frequencies correspond to those of sensitivity functions for M and P channels in LGN. Discussions: If neural activation in primary visual cortex propagates from the central visual field toward the peripheral filling-in target and if stronger activation propagates more quickly, a dynamic texture that induces stronger activation will greatly facilitate filling-in. For that reason, a dynamic texture having higher sensitive spatio-temporal frequency for M and P channels greatly facilitates filling-in.
Supported by JSPS (#15500332).