Abstract
It has previously been argued that first-order (FO) and second-order (SO) signals are kept independent up to and including the global-motion level (Edwards & Badcock, Vis Res 1994). That study found that contrast-modulated noise dots had no effect on the ability to extract a signal carried by a subset of luminance-defined dots, while luminance-modulated noise dots impaired the ability to extract a signal carried by contrast-modulated dots. It was argued that this asymmetry was due to the luminance-modulated dots being both a FO and SO stimulus. Here we address the question of whether the same degree of independence occurs in the pooling of one-dimensional local-motion signals, which are pooled via an intersection-of-constraints (IOC) process, rather than the vector-average process used for two-dimensional stimuli (Nishida, et.al. VSS, 2006). We used stimuli consisting of static Gabor patches that had moving carriers which were either luminance (FO) or contrast (SO) modulated. These stimuli were used to determine the effectiveness of one kind of noise (FO or SO) in disrupting the extraction of a global-motion signal carried by the other type of signal (SO or FO). Results were the same as the original study by Edwards and Badcock: SO noise did not mask FO signal extraction, but FO noise masked SO extraction. Additionally, given that the luminance-defined stimulus in this experiment was a pure FO stimulus, the asymmetry in the masking cannot be due to a SO signal in the luminance-defined (FO) stimulus. Taken together, these results suggest that FO and SO interactions are the same for both IOC and vector-average pooling.
Supported by ARC Centre of Excellence for Visual Science Grant CE0561903.