Neocognitron
Encyclopedia
The neocognitron is a hierarchical multilayered neural network
proposed by Professor Kunihiko Fukushima. It has been used for handwritten character recognition
and other pattern recognition
tasks.
The neocognitron is inspired by the model proposed by Hubel & Wiesel
in 1959. They found two types of cells in visual primary cortex called simple cell
and complex cell
, and also proposed a cascading model of these two type of cells.
The neocognitron is a natural extension of these cascading models. In the neocognitron, which consists of multiple types of cells the most important of which are called S-cells and C-cells, the local features are extracted by S-cells, and these features' deformation, such as local shifts, are tolerated by C-cells. Local features in the input are integrated gradually and classifying in the higher layers. The idea of local feature integration is in several other models such as LeNet and SIFT
model.
There are multiple kinds of neocognitron. For example, some types of neocognitron can detect multiple patterns in the same input by using backward signals to achieve selective attention.
Artificial neural network
An artificial neural network , usually called neural network , is a mathematical model or computational model that is inspired by the structure and/or functional aspects of biological neural networks. A neural network consists of an interconnected group of artificial neurons, and it processes...
proposed by Professor Kunihiko Fukushima. It has been used for handwritten character recognition
Handwriting recognition
Handwriting recognition is the ability of a computer to receive and interpret intelligible handwritten input from sources such as paper documents, photographs, touch-screens and other devices. The image of the written text may be sensed "off line" from a piece of paper by optical scanning or...
and other pattern recognition
Pattern recognition
In machine learning, pattern recognition is the assignment of some sort of output value to a given input value , according to some specific algorithm. An example of pattern recognition is classification, which attempts to assign each input value to one of a given set of classes...
tasks.
The neocognitron is inspired by the model proposed by Hubel & Wiesel
Torsten Wiesel
Torsten Nils Wiesel was a Swedish co-recipient with David H. Hubel of the 1981 Nobel Prize in Physiology or Medicine, for their discoveries concerning information processing in the visual system; the prize was shared with Roger W...
in 1959. They found two types of cells in visual primary cortex called simple cell
Simple cell
A simple cell in the primary visual cortex is a cell that responds primarily to oriented edges and gratings . These cells were discovered by Torsten Wiesel and David Hubel in the late 1950s ....
and complex cell
Complex cell
Complex cells can be found in the primary visual cortex , the secondary visual cortex , and Brodmann area 19 .Like a simple cell, a complex cell will respond primarily to oriented edges and gratings, however it has a degree of spatial invariance. This means that its receptive field cannot be...
, and also proposed a cascading model of these two type of cells.
The neocognitron is a natural extension of these cascading models. In the neocognitron, which consists of multiple types of cells the most important of which are called S-cells and C-cells, the local features are extracted by S-cells, and these features' deformation, such as local shifts, are tolerated by C-cells. Local features in the input are integrated gradually and classifying in the higher layers. The idea of local feature integration is in several other models such as LeNet and SIFT
Scale-invariant feature transform
Scale-invariant feature transform is an algorithm in computer vision to detect and describe local features in images. The algorithm was published by David Lowe in 1999....
model.
There are multiple kinds of neocognitron. For example, some types of neocognitron can detect multiple patterns in the same input by using backward signals to achieve selective attention.
External links
- Neocognitron on ScholarpediaScholarpediaScholarpedia is an English-language online wiki-based encyclopedia that uses the same MediaWiki software as Wikipedia, but has features more commonly associated with open-access online academic journals....
- NeoCognitron by Ing. Gabriel Minarik - application (C#) and video
- Neocognitron resources at Visiome Platform - includes MATLAB environment
- Beholder - a Neocognitron demo project
- Neocognitron on Learn artificial neural networks