Artificial intelligence · 1980

Neocognitron

Fukushima builds a layered vision network with local features and pooling.

Kunihiko Fukushima's neocognitron stacked alternating layers of feature detectors and shift-tolerant pooling, inspired by Hubel and Wiesel's visual cortex findings. It is the architectural ancestor of the convolutional network, lacking only backpropagation training.

See it in the living tree Wikipedia

Arcs above join it to what it stood on, arcs below to what it led to. The bar places this stretch on the tree's whole timeline.

What it built on 1

Perceptron 1958

What it led to 1

Convolutional neural network 1989

Related measurements

The dashed line marks when it first worked.

Training compute of notable AI models Source: Epoch AI, Notable AI Models (CC BY 4.0)

Computation per constant dollar, 1850-2006 Latest: 7.3B million units of computer power per 2006 $ (2003) Source: Nordhaus 2007, Two Centuries of Productivity Growth in Computing, J. Econ. Hist. 67(1), Table 6