Animate–Inanimate Object Categorization from Minimal Visual Information in the Human Brain, Human Behavior, and Deep Neural Networks

Abstract The distinction between animate and inanimate things is a main organizing principle of information in perception and cognition. Yet, animacy, as a visual property, has so far eluded operationalization. Which visual features are necessary and sufficient to discriminate between animate and inanimate objects? At which level of the visual hierarchy does the animate–inanimate distinction emerge? Here, we show that the categorical boundary between animate and inanimate objects is preserved even among images of objects that are made unrecognizable and only retain the low- and mid-level visual features of their natural version. In three experiments, healthy human adults viewed rapid sequences of images (6 Hz), in which an exemplar from another category (e.g., animate) was presented after every five exemplars from a given category (i.e., inanimate), at a rate of 1.2 Hz. Using frequency-tagging electroencephalography, we found significant neural responses at 1.2 Hz, indicating rapid and automatic detection of the periodic categorical change. Moreover, such effect was found—although increasingly weaker—for “impoverished” stimulus sets that retained only certain (high-, mid-, or low-level) features of the original colorful images (i.e., grayscale, texform, and phase-scramble images), and even if the images were unrecognizable, as tested with an object recognition task. Similar effects were found with deep neural networks presented with the same stimulus sets. Reliable categorization effects for dramatically impoverished and unrecognizable images, in humans' EEG and deep neural network data, demonstrate that the animate–inanimate distinction emerges early in the visual hierarchy and is remarkably resilient to the loss of visual information.