Multinomial Bayesian Learning for Modeling Classical and Nonclassical Receptive Field Properties
Multinomial Bayesian Learning for Modeling Classical and Nonclassical Receptive Field Properties
复制标题
DOI：
10.1162/neco_a_00310
复制
发表时间：
2012-08-01
影响因子：
2.9
通讯作者：
Hosoya, Haruo
中科院分区：
文献类型：
--
作者：
Hosoya, Haruo
我们研究了贝叶斯推理和自然图像学习在分层视觉系统中的相互作用，与早期视觉皮层的响应特性有关。我们特别关注的贝叶斯网络与多项式变量，可以表示离散的特征空间类似于超列结合minicolumns，执行稀疏的激活学习有效的表示，并解释分裂规范化。我们证明了使用基于采样的贝叶斯推理的最大似然学习产生了类似于V1简单细胞和V2细胞的经典感受野特性，而在训练的网络上进行的推理产生了非经典的上下文相关的响应特性，如交叉方向抑制和填充。与已知的生理特性的比较揭示了一些定性和定量的相似性。
We study the interplay of Bayesian inference and natural image learning in a hierarchical vision system, in relation to the response properties of early visual cortex. We particularly focus on a Bayesian network with multinomial variables that can represent discrete feature spaces similar to hypercolumns combining minicolumns, enforce sparsity of activation to learn efficient representations, and explain divisive normalization. We demonstrate that maximal-likelihood learning using sampling-based Bayesian inference gives rise to classical receptive field properties similar to V1 simple cells and V2 cells, while inference performed on the trained network yields nonclassical context-dependent response properties such as cross-orientation suppression and filling in. Comparison with known physiological properties reveals some qualitative and quantitative similarities.