Reverse engineering the origins of visual intelligence

AbstractIn recent years, researchers have made great strides developing a mechanistic understanding of object recognition in mature brains. Despite this progress, fundamental questions remain about the origins and development of object recognition. To what extent is the ‘initial state’ of object recognition innately constrained? What are the learning algorithms that transform the initial, naïve state into a mature state? Here, we describe a new experimental approach for studying the origins and development of object recognition, by performing parallel controlled-rearing experiments on newborn chicks and autonomous artificial agents. This approach can be used to isolate the core computational components that underlie visual intelligence.

Return to previous page