Visual perceptual learning (VPL), the training-induced improvement in visual tasks, has long been considered the product of neural plasticity at early and local stages of signal processing. However, recent evidence suggests that multiple networks and mechanisms, including stimulus- and task-specific plasticity, concur in generating VPL. Accordingly, early models of VPL, which characterized learning as being local and mostly involving early sensory areas, such as V1, have been updated to embrace these newfound complexities, acknowledging the involvement on parietal (i.e. intra-parietal sulcus) and frontal (i.e. dorsolateral prefrontal cortex) areas, in aspects concerning decision-making, feedback integration and task structure. However, evidence of multiple brain regions differentially involved in different aspects of learning is thus far mostly correlational, emerging from electrophysiological and neuroimaging techniques. To directly address these multiple components of VPL, we propose to use a causal neuromodulation technique, namely transcranial random noise stimulation, to selectively modulate the activity of different brain regions suggested to be involved in various aspects of learning. Specifically, we will target a region in the occipital cortex, which has been associated with stimulus-specific plasticity, and one in the parietal cortex, which has been associated with task-specific plasticity, in a between-subject design. Measures of transfer of learning to untrained stimuli and tasks will be used to evaluate the role of different regions and test for double dissociations between learning effects and stimulated area, shedding lights on learning mechanisms in the visual system. Evidence of dissociable mechanisms of learning can help refine current models of VPL and may help develop more effective visual training and rehabilitation protocols.