Sensory integration for space perception based on scalar learning rule
Takao Maeda, Susumu Tachi, Eimei Oyama · 2005
From a psychophysical viewpoint, the human sensory space does not completely coincide with physical space. The purpose of this study is to clarify why such a human perceptional space does not completely coincide with a physical one. Toward this end, we propose a learning rule and a neural network model using it. We call the learning rule scalar learning rule and name the model independent scalar learning elements summation model (ISLES model). The space discordance phenomena reflected in the model are similar to human ones reported in many psychophysical experiments. Therefore, the neural network model can be a good approximation to the physiological process of human space perceptions.