Graduate Student, Kuhlman-Lab, Carnegie Mellon University
Abstract: Significance: Using population decoding algorithms we discovered that attribute-invariance is represented at the earliest stages of cortical processing. This is interesting because invariance, a key aspect of object recognition that is typically studied as size-invariance or object rotation-invariance, is thought to be a property that emerges at higher cortical areas, rather than V1.
Here, we designed a novel paradigm to reveal that unexpectedly, invariant attributes can in fact be represented in V1. Thus, V1 neurons do more than merely edge-detection, as the texts books would have you think! Future studies will explore the role of top-down feedback and experience in developing this capacity.