Abstract
Performance on many low-level visual tasks improves with practice, with improvements usually limited to trained stimulus and task dimensions. However, in visually impaired populations, e.g., those with cortical blindness or amblyopia, visual training can improve performance on untrained stimuli and tasks (Huxlin et al, 2009; Huang et al, 2008; Zhou et al. 2006). Contrast sensitivity (CS) features prominently among these collaterally improved functions (Huxlin et al., 2009; Polat, 2009). Potential explanations for this include: (1) damaged visual systems possess greater-than-normal plasticity; (2) stimuli/tasks used for rehabilitation induce broad transfer of learning, and (3) CS improvements are a natural consequence of visual training. To test these proposed explanations, three groups of visually intact participants underwent baseline measurement of contrast sensitivity functions (CSFs) for motion direction and orientation discrimination in the periphery. Group 1 then trained on an orientation discrimination task (using 2 cycles/deg, static, non-flickering Gabors). Group 2 trained on a global direction discrimination task previously used to recover motion perception in cortically blind patients (Huxlin et al., 2009). Group 3 underwent no training. After 10 days, the battery of CS measurements was repeated. While both forms of training improved CS, Group 1 exhibited its largest enhancement for CSFs measured with static stimuli similar to those used during training, whereas Group 2's CS improvements occurred in functions measured with moving or flickering stimuli. Group 2 also exhibited enhancements over a broader ranger of spatial and temporal frequencies. Group 3 generally saw no significant CSF improvement from test to re-test. Thus, CS improvements appear to be a consequence of many, disparate forms of visual training, while retaining some degree of specificity for basic attributes of the trained stimulus. It remains to be determined whether CS improvements are critical for visual learning, or whether they are a simple by-product of repetitive visual training.
Meeting abstract presented at VSS 2014