Recognition of object position and shape by the spread pattern of a spatial spreading associative neural network
Kiyomi Nakamura, Akira Takarajima · Systems and Computers in Japan · 1997
Recently, there have been intensive studies of object position and shape recognition using an artificial neural network that resembles or mimics brain structure. In the brain, both spatial and object recognition systems are considered to work in parallel, and cooperatively. In the present study, a spatial spreading associative neural network that pays attention to the spatial recognition system in the brain is developed. The spatial spreading associative neural network achieves both shape-invariant positional recognition and position-invariant shape recognition of the object at the same time. The spatial spreading associative neural network spreads the input pattern by a spatial spreading neural layer. The spatial spreading neural layer has similar positional tuning characteristics to the directional discrimination neuron in the parietal cortex. The positionally spread pattern is then associated with the direction and shape memory neurons by generalized inverse learning. A population vector recognizes the position of an object in the input pattern irrespective of its shape. The shape memory neuron of the spatial spreading associative neural network recognizes the shape of the object in the input pattern irrespective of its position. The tuning width of the spatial spreading neural layer is critical for invariant recognition. © 1998 Scripta Technica, Syst Comp Jpn, 28(10): 9–19, 1997