Abstract
Perceptual learning generally transfers between contrast-defined motion and luminance-defined motion, but not vice-versa (Mather & West, 1993; Lu & Sperling, 1995; Scott-Samuel & Georgeson, 1999). Vaina & Chubb (2012), however, reported no transfer either way, suggesting entirely separate 1st - and 2nd-order motion mechanisms. Here, we report the results of perceptual learning transfer between the 1st and 2nd order mechanisms mediating fine, static orientation discriminations. Ten observers participated in the study. Four had participated in several experiments previously, and were highly practiced making orientation judgments with 4 cpd luminance modulated (LM) gratings (contrast = 0.1). We then trained them with contrast-modulated, binary noise gratings (CM), also 4 cpd. We initially tested them at the same orientation difference that yielded a d' of 1.5 with LM gratings (approximately ±0.5 deg), also at a modulation depth of 0.1. No observer could perform the task above chance. We had to revert to ± 45 deg and shape responses over a period of months. Thresholds were about 10 times higher than found with LM gratings. Another four observers were new to the lab, and were initially trained with the CM gratings. Again, we found we had to start at ±45 deg and shape down to a final difference threshold of about ±5.0 deg. There was no difference in the trajectories of the two groups, suggesting that there was no transfer of perceptual learning from the 1st order to the 2nd order mechanisms. We are in the process of testing for transfer of perceptual learning from the CM to the LM mechanisms. Preliminary evidence suggests that there may be some transfer, but it is not yet clear whether this will be significant. The remaining two observers, also new to the lab, were trained on LM gratings, to provide the control learning trajectory against which the latter transfer group will be compared.
Meeting abstract presented at VSS 2014