Energy function for learning invariance in multilayer perceptron

被引：11

作者：

Peng, HC ^{[1
]}

Sha, LF ^{[1
]}

Gan, Q ^{[1
]}

Wei, Y ^{[1
]}

机构：

[1] SE Univ, Dept Biomed Engn, Nanjing 210096, Peoples R China

来源：

ELECTRONICS LETTERS | 1998年 / 34卷 / 03期

关键词：

D O I：

10.1049/el:19980161

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

A new energy function is proposed for forming self-adapting ordered representations of input samples in a multilayer perceptron. Simulation results on unconstrained handwritten digit recognition give a better invariance extraction for this model than for several other models.

引用

页码：292 / 294

页数：3

共 5 条

[1] Learning Invariance from Transformation Sequences [J].

Foldiak, Peter .

NEURAL COMPUTATION, 1991, 3 (02) :194-200

[2] Emergence of simple-cell receptive field properties by learning a sparse code for natural images [J].

Olshausen, BA ;

Field, DJ .

NATURE, 1996, 381 (6583) :607-609

[3] A LEARNING RULE FOR EXTRACTING SPATIOTEMPORAL INVARIANCES [J].

STONE, J ;

BRAY, A .

NETWORK-COMPUTATION IN NEURAL SYSTEMS, 1995, 6 (03) :429-436

[4] Using spatio-temporal correlations to learn invariant object recognition [J].

Wallis, G .

NEURAL NETWORKS, 1996, 9 (09) :1513-1519

[5] Optimal, unsupervised learning in invariant object recognition [J].

Wallis, G ;

Baddeley, R .

NEURAL COMPUTATION, 1997, 9 (04) :883-894

← 1 →