Part Segmentation for Object Recognition

被引：23

作者：

Pentland, Alex ^{[1
]}

机构：

[1] MIT, Media Lab, Vis Sci Grp, Room E15-410,20 Ames St, Cambridge, MA 02139 USA

来源：

NEURAL COMPUTATION | 1989年 / 1卷 / 01期

基金：

美国国家科学基金会;

关键词：

D O I：

10.1162/neco.1989.1.1.82

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Visual object recognition is a difficult problem that has been solved by biological visual systems. An approach to object recognition is described in which the image is segmented into parts using two simple, biologically-plausible mechanisms: a filtering operation to produce a large set of potential object "parts," followed by a new type of network that searches among these part hypotheses to produce the simplest, most likely description of the image's part structure.

引用

页码：82 / 91

页数：10

共 10 条

[1] PARALLEL VISUAL COMPUTATION [J].

BALLARD, DH ;

HINTON, GE ;

SEJNOWSKI, TJ .

NATURE, 1983, 306 (5938) :21-26

[2]

BINFORD TO, 1971, P IEEE C SYST CONTR

[3]

HOFFMAN D, 1985, PIXELS PREDICATES

[4]

HOPFIELD JJ, 1985, BIOL CYBERN, V52, P141

[5] ON THE FOUNDATIONS OF RELAXATION LABELING PROCESSES [J].

HUMMEL, RA ;

ZUCKER, SW .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1983, 5 (03) :267-287

[6]

Leclerc Y., 1988, P DARPA IM UND WORKS, P365

[7] REPRESENTATION AND RECOGNITION OF SPATIAL-ORGANIZATION OF 3-DIMENSIONAL SHAPES [J].

MARR, D ;

NISHIHARA, HK .

PROCEEDINGS OF THE ROYAL SOCIETY SERIES B-BIOLOGICAL SCIENCES, 1978, 200 (1140) :269-294

[8]

Pentland A., 1988, 104 MIT MED LAB VIS

[9] COMPUTATIONAL VISION AND REGULARIZATION THEORY [J].

POGGIO, T ;

TORRE, V ;

KOCH, C .

NATURE, 1985, 317 (6035) :314-319

[10]

Rissanen J., 1983, ENCY STAT SCI, V5, P523

← 1 →