Object similarity affects the perceptual strategy underlying invariant visual object recognition in rats

摘要: In recent years, a number of studies have explored the possible use rats as models high-level visual functions. One central question at root such an investigation is to understand whether rat object vision relies on processing shape features or, rather, lower-order image properties (e.g., overall brightness). study, we shown that are capable extracting multiple diagnostic its identity, least when those are, structure-wise, distinct enough be parsed by system. present assessed impact structure perceptual strategy. We trained discriminate between two structurally similar objects, and compared their recognition strategies with reported in our previous study. found that, under conditions lower stimulus discriminability, discrimination strategy becomes more view-dependent subject-dependent. Rats were still able recognize target way was largely tolerant (i.e., invariant) transformation; however, larger structural pixel-wise similarity affected objects processed. Compared findings patterns were: (i) smaller scattered; (ii) only partially preserved across views; (iii) reproducible rats. On other hand, adopt multi-featural make part optimal discriminatory information afforded objects. Our suggest humans, invariant can flexibly rely either view-invariant representations distinctive or view-specific representations, acquired through learning.