Dimension reduction, generalized deformable models and the development of ocularity and orientation

被引:14
作者
Yuille, AL
Kolodny, JA
Lee, CW
机构
[1] Division of Applied Sciences, Harvard University, Cambridge
关键词
developmental models; Hebb's rule;
D O I
10.1016/0893-6080(95)00085-2
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
It has been proposed that the spatial organization of cells in visual cortex satisfies an abstract optimization criteria based on a smooth mapping criterion. We apply this criteria to the development of ocularity and show that it can be expressed in terms of a generalized deformable model. From this model we can derive two alternative algorithms, one being an elastic net and the other being of Hebbian form. This shows a surprising mathematical relation between the models of Goodhill and Willshaw (1990, Network, 1, 41-59) and Miller et al. (1989, Science, 245, 605-615), even though their motivations are very different. Applying the same techniques to the development of orientation shows mathematically that a variant of the dimension reduction criterion (Durbin & Mitchison, 1990, Nature, 343, 644-647) can be closely approximated by an elastic net.
引用
收藏
页码:309 / 319
页数:11
相关论文
共 32 条
[1]   VOLTAGE-SENSITIVE DYES REVEAL A MODULAR ORGANIZATION IN MONKEY STRIATE CORTEX [J].
BLASDEL, GG ;
SALAMA, G .
NATURE, 1986, 321 (6070) :579-585
[2]  
BLASDEL GG, 1989, ANNU REV PHYSIOL, V51, P561, DOI 10.1146/annurev.physiol.51.1.561
[3]   ORIENTATION SELECTIVITY, PREFERENCE, AND CONTINUITY IN MONKEY STRIATE CORTEX [J].
BLASDEL, GG .
JOURNAL OF NEUROSCIENCE, 1992, 12 (08) :3139-3161
[5]  
BLASDEL GG, 1988, SENSORY SYSTEMS
[6]   AN ANALOG APPROACH TO THE TRAVELING SALESMAN PROBLEM USING AN ELASTIC NET METHOD [J].
DURBIN, R ;
WILLSHAW, D .
NATURE, 1987, 326 (6114) :689-691
[7]   A DIMENSION REDUCTION FRAMEWORK FOR UNDERSTANDING CORTICAL MAPS [J].
DURBIN, R ;
MITCHISON, G .
NATURE, 1990, 343 (6259) :644-647
[8]   An Analysis of the Elastic Net Approach to the Traveling Salesman Problem [J].
Durbin, Richard ;
Szeliski, Richard ;
Yuille, Alan .
NEURAL COMPUTATION, 1989, 1 (03) :348-358
[9]  
ERWIN E, 1994, UIUCBITB9408 U ILL U
[10]  
GEIGER D, 1989, 1114 MIT ART INT LAB