Abstract
Because the complexity of visual features for which neurons are tuned increases from early to late stages of the ventral visual stream, the working hypothesis is that object recognition, categorization, and novelty/memory perception are primarily mediated by higher visual areas. However, observers can categorize images as objects or animals or as big or small even when the images have been filtered to preserve some low-level features but are rendered unidentifiable (‘texforms’, Long, Chen & Konkle, 2018). This observation implies that even early visual cortex, in which neurons respond to simple stimulus features, may encode signals that are of use in seemingly high-level visual functions. We tested this hypothesis by recording from populations of neurons in early and mid-level visual cortical areas while rhesus monkeys viewed texforms and their unaltered source stimuli (simultaneous recordings from areas V1 and V4 in one animal and from V4 in a second). Using recordings from a few dozen neurons, we could decode the size and animacy of both original images and texforms. We also observed repetition suppression (reduced responses to familiar rather than novel stimuli) to both stimulus categories, even when the second presentation of a stimulus occurred tens of seconds after the first presentation with multiple other stimuli and eye movements in between. We compared category encoding and repetition suppression to the ability of human observers to identify and categorize the texforms that originated from different stimuli, and found that the humans’ categorization performance is related to the ability to decode size and animacy from V1 and V4. Our results demonstrate that neuronal populations early in the visual hierarchy contain signals useful for higher-level object perception and memory, and suggest that these responses to simple stimulus features show preliminary untangling of higher-level distinctions.