Abstract
Both recent visual history and surrounding visual context modulate discriminative accuracy. Their respective effects upon discrimination performance depend similarly upon the angular difference between the inducing (adapting or surround) and test stimulus. We have recently shown that repeatedly practising a fine discrimination task while in an adapted state completely reconfigures the characteristic effects of adaptation. Here we ask whether learning produces similar effects on surround modulation of discrimination performance. We asked a group of human subjects to report whether the orientation of a sinusoidal grating was tilted clockwise or counter-clockwise of vertical. Before and after training, we measured the baseline discrimination threshold and change in this estimate caused by the presence of a vertical annular surround grating (0°) or two surround gratings oriented symmetrically (± 10 to ± 50°) around vertical. During training, the same group of subjects repeatedly practiced (over 8-10 daily training sessions) discriminating the orientation of the central grating while it was surrounded by two gratings that induced the largest threshold elevation before training. We found that practicing orientation discrimination with surround stimuli present produced significant improvements in discriminative accuracy. Like the effects of learning in an adapted state, these large improvements in the trained surround condition had consequences for untrained surround conditions. Surround conditions that prior to learning had little effect induced a significant cost for discriminative accuracy after learning. Our results suggest that the behavioural signatures of surround modulation are flexible, which has implications for the sort of inferences we can make about surround-induced changes to performance.
Meeting abstract presented at VSS 2013