Journal Home Online First Current Issue Archive For Authors Journal Information 中文版

Strategic Study of CAE >> 2005, Volume 7, Issue 2

Improved Matching Algorithms for Linear Face Class Model

Institute of Artificial Intelligence and Robotics, Xi'an Jiaotong University, Xi'an 710049, China

Funding project:国家自然科学创新研究群体基金资助项目(60024301);国家自然科学基金资助项目(60205001);河南省重大科技基金资助项目(0222020400) Received: 2004-01-08 Revised: 2004-03-08 Available online: 2005-02-20

Next Previous

Abstract

An advanced matching technique for linear face class model is proposed, which can solve the problem of detailed controlling and robust iteration for the realistic facial modeling. A new method——Dynamic Gaussian Pyramid Analysis (DGPA), which combines Non-Uniform Sampling (NUS) method and Multi-Resolution Analysis, is presented. Integrating the PS Sampling and the Cluster Random Sampling, the distribution of the sampled points in each level images of the Gaussian pyramid is adjusted dynamically. In coarse-to-fine scheme, the minimization algorithm is used to compute the near global optimal solution that may fit to yield accurate model matching. Dynamic adjusting the boundary of the sampling cluster area and the resampling ratio, the detailed representations are effectively controlled, and the model creation is quite robust. An improved Stochastic Gradient Descent (SGD) algorithm based on the Correlative Disturbance (CD) and Adaptive Learning Rate (ALR) is exploited to accelerate iteration convergence and compute valid model parameters. With the examples of MPI Caucasian Face and AI&R Asian Face databases, experimental results in subjective evaluation and objective evaluation demonstrate the advanced model matching technique.

Figures

图1

图10

图11

图12

图13

图2

图3

图4

图5

图6

图7

图8

图9

References

[ 1 ] 郑南宁, 付 昀, 张 婷, 卓 峰.人脸的表情与年龄变换和非完整信息的重构技术 (上) [J].电子学报, 2003, 31 (12A) :1955~1962 link1

[ 2 ] 付 昀, 郑南宁, 张 婷.人脸的表情与年龄变换和非完整信息的重构技术 (下) [J].电子学报, 2003, 31 (12A) :1963~1970 link1

[ 3 ] 张 强.基于稠密特征对应的人脸图像表达及人脸属性变换[D].西安:西安交通大学, 2003

[ 4 ] 陈 洪.人脸图像自动分析与绘制的统计学习方法[D].西安:西安交通大学, 2002

[ 5 ] 卓 峰, 徐维朴, 张 强, 等.基于2D样本的人脸图像视点变换[J].计算机应用, 2003, 23 (增刊) :74~76 link1

[ 6 ] BlanzV , VetterT .Amorphablemodelforthesynthesisof3Dfaces[A].ComputerGraphicsAnnualConferenceSeries, SIGGRAPH 99[C ].ACMSIGGRAPH , LosAngeles, California, 1999.187~194

[ 7 ] VetterT .Synthesisofnovelviewsfromasinglefaceimage[J].InternationalJournalofComputerVision, 1998, 28 (2) :103~116 link1

[ 8 ] TimothyFCootes, GarethJEdwards, ChristopherJTaylor.Activeappearancemodels[J ].IEEETransactionsonPatternAnalysisandMachineIntelligence, 2001, 23 (6) :681~685

[ 9 ] UllmanS , BasriR .Recognitionbylinearcombinationsofmodels[J].IEEETransactionsonPatternAnalysisandMachineIntelligence, 1991, 13:992~1006

[10] VetterT , PoggioT .Linearobjectclassesandimagesynthesisfromasingleexampleimage[J].IEEETransactionsonPatternAnalysisandMachineIntelligence, 1997, 19 (7) :733~742

[11] JonesMJ , PoggioT .Model basedmatchingoflinedrawingsbylinearcombinationsofexamples[A ].IEEEInternationalConferenceonComputerVision (ICCV’95) [C], BostonMassachusetts, 1995.532~536

[12] JonesMJ, PoggioT .Multidimensionalmorphablemodels[A ].IEEEInternationalConferenceonComputerVision (ICCV’98) [C], IEEEComputerSociety, Bombay, 1998.683~688

[13] VetterT , JonesMJ, PoggioT .Abootstrappingalgorithmforlearninglinearmodelsofobjectclasses[A ].IEEEInternationalConferenceonComputerVisionandPatternRecognition (CVPR’97) [C], USA :IEEEComputerSocietyPress, PuertoRico, 1997.40~47

[14] BeymerD , PoggioT .Imagerepresentationforvisuallearning[J].Science, 1996, 272:1905~1909

[15] BeymerD .Vectorizingfaceimagesbyinterleavingshapeandtexturecomputations[R].A .I.Memo1537, ArtificialIntelligenceLaboratory, Massa chusetts:MIT , 1995

[16] JonesMJ, PoggioT .Model basedmatchingbylinearcombinationsofexamples[R].AIMemoNo1583, ArtificialIntelligenceLaboratory, Massachusetts:MIT , 1996

[17] BeymerD , ShashuaA , PoggioT .Examplebasedimageanalysisandsynthesis[R].AIMemoNo1431, ArtificialIntelligenceLaboratory, Massachusetts:MIT , 1993

[18] TorreFDela, BlackMJ.Robustprincipalcomponentanalysisforcomputervision[A].IEEEInternationalConferenceonComputerVision (ICCV’01) [C ], IEEEComputerSociety, Vancouver, Canada, 2001.362~369

[19] LeemonCBairdIII .Reinforcementlearningthroughgradientdescent[D].Pittsburgh:SchoolofComputerScienceatCarnegieMellonUniversity, 1999

[20] GuMG , KongFH .AstochasticapproximationalgorthmwithMarkovChainMonteCarloMethodforincompletedataestimationproblems[A].ProceedingsofNationalAcademyofSciences[C], 1998, 95:7270~7274

[21] PeterJBurt, EdwardII .Adelson.TheLaplacian  pyramidasacompactimagecode[J].IEEETransactionsonCommunications, 1983, COM -31 (4) :532~540

[22] BeierT , NeelyS .Feature basedimagemetamorphosis[J].ComputerGraphics, 1992, 26 (2) :35~42

[23] RichardODuda, PeterEHart, DavidGStork.PatternClassification, SecondEdition[M ].WileyInterscience, 2000

[24] TomMMitchell.MachineLearning[M ].USA :MCGRWW HILL , 1997

[25] 金勇进, 蒋 妍, 李序颖.抽样技术[M].北京:中国人民大学出版社, 2002 link1

[26] SeizeSM .Viewmorphing[A].ComputerGraphicsAnnualConferenceSeries, SIGGRAPH’96[C], ACMSIGGRAPH , NewOrleans, Louisiana, August, 1996.21~30

[27] WolbergG .Imagemorphingsurvey[J].TheVisualComputer, 1999, 14:360~372 link1

Related Research