Kate Walsh¹ (kate.walsh@colorado.edu), Rosemary Cowell¹, David Huber¹; ¹University of Colorado Boulder

Monocular perceptual learning transfers to the untrained eye (e.g., Lu et al., 2005), indicating learning within binocular brain regions. This implies an absence of learning within monocular brain regions, but it is not clear if this finding applies to between-eye, dichoptic learning. To address this question, we developed a novel inter-ocular perceptual learning paradigm in which perception of a Kanizsa stimulus (an illusory rectangle) can only be achieved by combining information across the eyes. This was achieved with a pair of pacman-shaped objects with one presented to one eye and the other presented to the other eye. This between-eye training condition was compared to within-eye training. Over six days of training, observers (N=14) judged the alignment of a monocularly-presented stimulus pair (“within-eye”) trained in one visual quadrant and a dichoptically-presented pair (“between-eye”) trained in the diagonally opposite quadrant. Over two transfer sessions, observers judged alignment of stimulus pairs, trained and untrained, across all quadrants to investigate the eye- and location-specificity of monocular and inter-ocular learning. Psychometric functions exhibited an above zero flat portion as a function of increasing misalignment, prior to the first increase in accuracy, indicating a bias to respond ‘misaligned’. A novel signal-detection based function was developed to capture this response bias, thus allowing measurement of the point of initial rise, which we termed the ‘limen’ (i.e., the smallest noticeable misalignment). Although performance was generally better for trained quadrants and trained stimuli, this learning did not reliably map onto specific parameters of the psychometric curve, with the exception of the limen, which revealed a hyper-specific training effect for the particular trained eye-combination in the quadrant that experienced between-eye training. This specificity supports the conclusion that between-eye learning occurred within monocular brain regions.