Three-dimensional object recognition using a recurrent attractor neural network

Author

Yoon, Richard S. ; Borrett, Don S. ; Kwan, Hon C.

Author_Institution

Dept. of Physiol., Toronto Univ., Ont., Canada

Volume

1

fYear

1995

fDate

20-25 Sep 1995

Firstpage

379

Abstract

Recognition of 3D objects on the basis of a 2D perspective view is performed effortlessly by many higher nervous systems, yet is not easily duplicated by machines. The present research presents a nonlinear dynamical approach to object recognition implemented by recurrent neural networks. Specifically, training orbits composed of coherent image sequences of distinct objects were used to partition the network phase space into appropriate basins of attraction. After training, network relaxation to appropriate attractor states constituted the process of recognition

Keywords

backpropagation; image recognition; image sequences; medical image processing; nonlinear dynamical systems; object recognition; recurrent neural nets; 2D perspective view; 3D objects; attractor states; basins of attraction; coherent image sequences; distinct objects; network phase space; network relaxation; neuroanatomical data; neurophysiological data; nonlinear dynamical approach; recurrent attractor neural network; recurrent neural network; three-dimensional object recognition; training orbits; wire frame objects; Biological neural networks; Educational institutions; Image recognition; Nervous system; Neural networks; Object recognition; Orbits; Physiology; Recurrent neural networks; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Engineering in Medicine and Biology Society, 1995., IEEE 17th Annual Conference

Conference_Location

Montreal, Que.

Print_ISBN

0-7803-2475-7

Type

conf

DOI

10.1109/IEMBS.1995.575159

Filename

575159