Abstract:

Purpose: Humans can learn to abstract and conceptualize the shared visual features defining an object category in object learning. Therefore, learning is generalizable to transformations of familiar objects and even to new objects that di?er in other physical properties. In contrast, visual perceptual learning (VPL), improvement in discriminating ?ne di?erences of a basic visual feature through training, is commonly regarded as speci?c and low-level learning due to its speci?city to the trained stimulus conditions. Such location and orientation speci?city is taken as evidence for neural plasticity in V1 or improved readout of V1 signals. However, new training methods have shown complete VPL transfer across stimulus locations and orientations, suggesting the involvement of high-level cognitive processes. Here we report that VPL bears similar properties of object learning.
Methods & Results:  Speci?cally, we trained human observers to discriminate the explicit orientation of luminance gratings encoded by V1 neurons; or the implicit orientation of bilaterally symmetric dot patterns encoded by higher cortical areas. We compared the mutual transfer of learning between these two distinct stimuli. Learning in orientation discrimination of the symmetric dot patterns transferred completely to the gratings. In contrast, learning in orientation discrimination of the gratings transferred only partially to the dot patterns; but subsequent exposure to the same-oriented dot patterns in an irrelevant task (a Training-Plus-Exposure technique) further markedly reduced orientation thresholds for the dot patterns, achieving a complete learning transfer from the gratings to the dot patterns. Similarly, motion direction discrimination learning is transferable between the ?rst- and second-order motion signals, replicating the same patterns of learning transfer.
Conclusion: These results suggest that perceptual learning can take place at a conceptual level and generalize to stimuli with different physical properties. Our findings thus reconcile perceptual and object learning into a uni?ed framework.

Key words: perceptual learning, transfer, orientation, motion direction