Dissertation > Excellent graduate degree dissertation topics show

Facial expression recognition research based on manifold learning

Author: CaiLinBo
Tutor: YingZiLu
School: Wuyi University
Course: Signal and Information Processing
Keywords: Expression Recognition Manifold learning Contourlet Transform Local binary pattern Locally Linear Embedding Laplace feature map
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 171
Quote: 0
Read: Download Dissertation

Abstract


The facial expression is an important body language, in people's daily lives, only 7% of the information through the language passed, 55% of the information through facial expressions passed. The facial expression recognition is the use of computer image feature information extracted facial expression, expression image, depending on the characteristics attributed to seven different expression category, it allows the computer to be able to infer the mental state based on the results of the expression of image classification, in order to achieve the natural interaction between man and machine. Although facial expression recognition technology has made a lot of progress, but in real life, light, gesture, noise, cover, and other factors, to achieve large-scale application still need further study. This paper analyzes the facial expression recognition technology research status of computer facial expression recognition of a number of issues were discussed, focusing on the application of manifold learning method in facial expression recognition, and carried out a series of facial expression recognition experiments . Thesis work consisted primarily of the following aspects: 1. Convection-shaped learning nonlinear dimensionality reduction methods made key introduction, detailing several common manifold learning, such as: isometric mapping, local linear embedding, Lapu Las feature map, Hessian Locally Linear Embedding and local tangent space alignment and analysis of its main advantages and disadvantages. 2. Facial expression recognition method based on Contourlet transform and local linear embedding. The Contourlet transform theory are described in detail and applied to facial expression feature extraction to generate a multi-resolution, multi-scale facial expression characteristics. LLE algorithm for feature dimensionality reduction, experiments were carried out on the JAFFE database and Cohn-Kanada database with the Wavelet LLE SVM and PCA SVM were compared with the proposed Contourlet LLE non-specific facial expression recognition in JAFFE database and Cohn-Kanada database on the highest recognition rate can reach 63.81% and 69.1%, respectively, higher than the above two methods. (3) the original LBP operator, multi-resolution and rotation invariant LBP operator to uniform LBP operator was introduced by LBP operator, analyzes their advantages and disadvantages. Focus described in a uniform pattern LBP operator on facial expression image feature extraction. 4. Laplace feature map facial expression recognition method based on local binary pattern. Called \A large number of non-specific facial expression recognition experiments on JAFFE database and Cohn-Kanada database LBP operator parameters (P, R) and LBP features image grid divided on the experimental results, the LE algorithm compared with the linear dimensionality reduction methods PCA and LDA, the LE algorithm proposed in this paper LBP on the JAFFE database and Cohn-Kanada database recognition rate reached 70.48% and 70.95% respectively, were higher than the LBP PCA and LBP lda, the highest recognition rate to verify the validity of the LE algorithm.

Related Dissertations

  1. The Classification of High Dimsnsion Flew Field Based on Manifold Learning,V231.3
  2. Research on Temporal Information Recognition and Normalization,TP391.1
  3. Feature Extraction, Selection and Combination in Lipreading,TP391.41
  4. Study of Data Reduction Technique Based on Manifold Learning,TP311.13
  5. Research on Predicting Intrinsic Disorder Protein Structure Based on Supervision Manifold Learning Algorithm,Q51
  6. Alexithymia individual's facial expression recognition and attentional bias experimental study,B849
  7. Primary emotions and emotional expression recognition characteristics and relationship between the amount of words to master,B844.2
  8. Research on Contourlet Transform and Its Application on Image Processing,TP391.41
  9. The Research of Image Matching Method Based on Feature Descriptor,TP391.41
  10. Facial expression recognition based on sparse representation residuals fusion,TP391.41
  11. Research of Distributed Video Coding Based on Compressed Sensing,TN919.81
  12. Research on Learning-Based Super-Resolution and SR System Implementation,TP391.41
  13. Face recognition algorithm based on feature fusion research,TP391.41
  14. EEG -based emotion recognition,TP391.4
  15. Hash based on structured sparse spectrum image indexing algorithm,TP391.41
  16. Based on the robust non-contact acquisition palmprint recognition,TP391.41
  17. Based on compressed sensing reconstruction algorithm,TN911.7
  18. Visible and infrared image fusion algorithm,TP391.41
  19. Time series visualization of human motion and multi- index method,TP391.41
  20. Based on multi- location basis ISOMAP Algorithm Research,TP181
  21. Image fusion method based on variational partial differential equations,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile