初期視覚野におけるBorder-ownership信号の確率論的解釈

概要

論文の詳細を見る
Several recent model studies of visual cortex have used Bayesian networks and their belief propagation algorithms, and successfiully explained various physiological properties. This paper shows that a similar model can also explain another property called border-ownership, one of contextual effects in early visual cortex reported by Zhou et al. We show that border-ownership signals can be interpreted as posterior joint probabilities of a low-level edge property and a high-level figure property, and can readily be found in a typical hierarchical Bayesian network mimicking early visual cortex, under certain conditions. We also present the result of a computer simulation that model neurons in our Bayesian network can reproduce response properties qualitatively similar to physiological data.
社団法人電子情報通信学会の論文
2010-03-02