Automatic construction of active appearance models as an image coding problem

被引：64

作者：

Baker, S ^{[1
]}

Matthews, I ^{[1
]}

Schneider, J ^{[1
]}

机构：

[1] Carnegie Mellon Univ, Inst Robot, Pittsburgh, PA 15213 USA

来源：

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE | 2004年 / 26卷 / 10期

关键词：

Active Appearance Models; automatic construction; unsupervised learning; image coding; inverse compositional image alignment; quadratic smoothness priors;

D O I：

10.1109/TPAMI.2004.77

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The automatic construction of Active Appearance Models (AAMs) is usually posed as finding the location of the base mesh vertices in the input training images. In this paper, we repose the problem as an energy-minimizing image coding problem and propose an efficient gradient-descent algorithm to solve it.

引用

页码：1380 / 1384

页数：5

共 23 条

[1]

[Anonymous], P EUR C COMP VIS ECC

[2]

[Anonymous], NEURAL COMPUT, DOI DOI 10.1162/NECO.1995.7.5.889

[3]

BAKER S, 2004, CMURITR0414

[4] A morphable model for the synthesis of 3D faces [J].

Blanz, V ;

Vetter, T .

SIGGRAPH 99 CONFERENCE PROCEEDINGS, 1999, :187-194

[5]

Brand M, 2001, PROC CVPR IEEE, P456

[6] Active appearance models [J].

Cootes, TF ;

Edwards, GJ ;

Taylor, CJ .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2001, 23 (06) :681-685

[7]

Davies RH, 2003, LECT NOTES COMPUT SC, V2732, P38

[8] A minimum description length approach to statistical shape modeling [J].

Davies, RH ;

Twining, CJ ;

Cootes, TF ;

Waterton, JC ;

Taylor, CJ .

IEEE TRANSACTIONS ON MEDICAL IMAGING, 2002, 21 (05) :525-537

[9] Automatic learning of appearance face models [J].

De la Torre, F .

IEEE ICCV WORKSHOP ON RECOGNITION, ANALYSIS AND TRACKING OF FACES AND GESTURES IN REAL-TIME SYSTEMS, PROCEEDINGS, 2001, :32-39

[10]

FREY B, 1999, P IEEE C COMP VIS PA, P1416

← 1 2 3 →