[1]张一鸣,欧宗瑛,王 虹.基于C均值K近邻算法的面部表情识别[J].智能系统学报,2008,3(1):57-61.
ZHANG Yi-ming,OU Zong-ying,WANG Hong.Facial expression recognition based on Cmeans and K-nearest neighbor algorithms[J].CAAI Transactions on Intelligent Systems,2008,3(1):57-61.
点击复制
ZHANG Yi-ming,OU Zong-ying,WANG Hong.Facial expression recognition based on Cmeans and K-nearest neighbor algorithms[J].CAAI Transactions on Intelligent Systems,2008,3(1):57-61.
基于C均值K近邻算法的面部表情识别
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
3
期数:
2008年第1期
页码:
57-61
栏目:
学术论文—机器感知与模式识别
出版日期:
2008-02-25
- Title:
- Facial expression recognition based on Cmeans and K-nearest neighbor algorithms
- 文章编号:
- 1673-4785(2008)01-0057-05
- 关键词:
- 面部表情识别; C均值聚类; K近邻; Gabor小波; Fisherface判别分析
- Keywords:
- facial expression recognition; Cmeans clusterin g; K-nearest neighbor; Gabor wavelet; fisherface discriminant analysis
- 分类号:
- TP39141
- 文献标志码:
- A
- 摘要:
- 随着人工智能与模式识别技术的不断发展，面部表情识别在智能人机交互中发挥着越来越重要的作用.通过对人的面部表情分类的研究，提出了一种使用C均值聚类、K近邻算法的面部表情分类方法.对参加训练的表情图像先进行Gabor小波变换，然后使用Fisherface判别分析方法进行变换，求得特征空间.再将已进行Gabor变换的标准表情图像投影到特征空间，进行C均值聚类得到子类表情模板.对于一幅待识别的表情图像，使用K近邻算法与子类表情模板比较，将该表情图像分类.使用该方法，在公开的日本女人表情人脸库上实测达到了95 8%的识别率.
- Abstract:
- With the rapid development of artificial intelligence and pattern rec ognition, facial expression recognition plays an important role in intelligent humanmachine interaction. In this paper, a facial expression classific a tion method is presented which uses a Cmeans and Knearest neighbor a lgorithm a s the basis of analysis for the classification of facial expressions. First the images to be analyzed are transformed with Gabor wavelets, and then Fisherface d i scriminate analysis is performed to generate a feature space. Next, the images w hich were transformed with Gabor wavelets are projected into the feature space a n d Cmeans clustering performed on the projected images to generate subexp ress io n templates. Finally, the type of expression is identified by comparing the inpu t expression images with the subexpression templates by using a Knearest nei gh bor algorithm. Experiments on the public Japanese female facial expression datab as e show that the method proposed in this paper can achieve a 95.8% recognition rate.
- 参考文献/References:
-
［1］EKMAN P, FRIESEN W V. Facial action coding system: a technique for t he measurement of facial movement ［M］. Palo Alto, CA:Consulting Psychologists Press,1978.
［2］MASE K. Recognition of facial expression from optical flow ［J］. IEICE Tr ans E,1991,74(10)：34743483.
［3］YACOOB Y, DAVIS L. Recognizing human facial expressions from long image se quences using optical flow ［J］. IEEE Trans on PAMI,1996，18(6)：636642.
［4］COTTRELL G，METCALFE J. Face, gender and emotion recognition using Holons ［C］// Advances in Neural Information Processing Systems. Denver, USA, 1990,3: 564571.
［5］PADGETT C, COTTRELL G. Representing face images for emotion classificatio n［C］// Advances in Neural Informati on Processing Systems. Cambridge: MIT Press,1997.
［6］金 一, 阮秋琦.一种局部加权的二维主成分分析算法及其在人脸识别中的应用［J］.智能系统学报,2007,2(3):2529.
JIN Yi,RUAN Qiuqi.A part weighted twodimensional PCA for face recognition［J］ .CAAI Transactions on Intelligent Systems, 2007,2(3):2529.
［7］DONATO G,STEWART M B, HAGER J C, et al. Classifying facial acti ons ［J］. IEEE Trans on PAMI, 1999, 21(10): 974989.
［8］DAUGMAN J G. Complete discrete 2D Gabor transform by neural networks for image analysis and compression ［J］. IEEE Trans on ASSP，1998, 36(7): 1169117 9 .
［9］BUCIU I, KOTROPOULOS C.ICA and Gabor representation for facial ex pression recognition［C］// Proceedings of IEEE ICIP. Barcelona, Spain, 2003.
［10］PENEV P S, ATICK J J. Local feature analysis: a general statistica l theor y for object representation ［J］. Network: Computation in Neural Systems, 1996 ，7(3)：477500.
［11］CALDER A J, BURTON A M, MILLER P, et al. A principal component analysis o f facial expressions ［J］.Vision Research, 2001,41(9): 11791208.
［12］SEYEDARABI H, AGHAGOLZADEH A, KHAN MOHAMMADI S. Recognition of six basic f acial expressions by featurepoints tracking using RBF neural network and fuzzy inference system［C］// Proceedings of 2004 IEEE International Conference on Mu ltimedia and Expo. Taipei, China, 2004.
［13］MA L，KHORASANI K. Facial expression recognition using constructive feedf orward neural networks ［J］. IEEE Trans on SMCPart B，2004，34(3)：15881595 .
［14］LADES M，JAN C. Distortion invariant object recognition in the dynamic li nk architecture ［J］. IEEE Trans on Computer，1993, 42(3)：300311.
［15］边肇祺，张学工.模式识别［M］.北京：清华大学出版社，2000.
［16］LYONS M J，BUDYNEK J，AKAMATSU S. Automatic classification of single faci al images ［J］. IEEE Trans on PAMI，1999，21(12)：13571362.
备注/Memo
收稿日期：2007-02-10.
基金项目：大连理工大学与中科院沈阳自动化研究所联合探索基金资助项目(DUTSIA 2006).
作者简介：
张一鸣，女，1981年生，硕士研究生，主要研究方向为模式识别.
欧宗瑛，男，1936年生，教授，博士生导师，主要研究方向为计算机辅助设计、计算机图形学和图像处理.获教委和机械部科技三等奖各一项，辽宁省科技一等奖一项.参与主编的机械设计手册和机电类规划教材分别获1995全国科技图书二等奖和2002全国优秀教材二等奖. 发表的学术论文被SCI检索12篇，被EI检索95篇.
王 虹，男，1964年生，博士研究生，主要研究方向为模式识别、图像处理
通讯作者：欧宗瑛.E-mail:ouzyg@dlut.edu.cn.
更新日期/Last Update:
2009-05-10