Title :
Artificial neural networks which can see geometric illusions in human vision
Author :
Chao, Jinhui ; Kishigami, Tohru ; Minowa, Kenji ; Tsujii, Shigeo
Author_Institution :
Dept. of Electr. & Electron. Eng., Chuo Univ., Tokyo, Japan
Abstract :
A new physiological model is proposed for geometrical illusion phenomena in human vision and is implemented by artificial neural networks. According to the model, illusion is a result of nonuniform bending in visual space whose Riemannian metric tensor is determined by lateral excitatory-inhibitory dynamics of the retina neurons.
Keywords :
neural nets; neurophysiology; physiological models; visual perception; Riemannian metric tensor; geometric illusions; human vision; lateral excitatory-inhibitory dynamics; neural networks; physiological model; retina neurons; visual space; Artificial neural networks; Chaos; Humans; Information processing; Intelligent networks; Neurons; Psychology; Solid modeling; Space technology; Tensile stress;
Conference_Titel :
Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on
Print_ISBN :
0-7803-1421-2
DOI :
10.1109/IJCNN.1993.714165
