Dissertation > Excellent graduate degree dissertation topics show

Gabor-based Manifold Learning for Face Recognition

Author: ZhuGuanWei
Tutor: LiHong
School: Huazhong University of Science and Technology
Course: Computational Mathematics
Keywords: Manifold Learning Gabor Filter Feature Extraction Locality Sensitive Discriminant Analysis Face Recognition Local Tangent Space Alignment Supervised Learning Unsupervised Learning
CLC: TP391.41
Type: Master's thesis
Year: 2009
Downloads: 73
Quote: 0
Read: Download Dissertation

Abstract


Recently, a new kind of dimensionality reduction algorithm, named manifold learning, has drawn much attention. Manifold learning theory assumes that high-dimensional data lie on a low-dimensional submanifold, its algorithms attempt to embed the original data into a submanifold by preserving the local geometry structure. In this paper, we introduce some representative algorithms include Isomap, LLE, LE, LTSA, LSDA and so on, and then stress the application of these algorithms in the field of face recognition.Face recognition is an important area of computer pattern recognition research. While humans are highly easy in recognizing faces, this task remains a significant challenge for machines. In the field of face recognition, the Gabor wavelet, whose kernels are similar to the 2D receptive field profiles of the mammalian cortical simple cells, can achieve the minimum of the Heisenberg uncertainty principle, thus has the best time-frequency resolution. Gabor filter has been successfully used in featrue extaction applied in machine vision, image processing and so on. This paper presents a novel Gabor-based manifold learnning method for face recogniton. The main tasks are:Firstly, Because of the Muti-resolution characteristic of the Gabor filter, We first apply a series of Gabor filters to face image to extract the image features from different direction and scale. Then the processing of the features such as sampling, arangement can be done to get a feature vector. As we all know, low dimensionality is especially important for learning. The dimensionality of the Gabor vector is very high, thus we apply manifold learning algriothms to reduce these high-dimensional features, then recognition work can be done in the low-dimensional subspace. In this article, we inctroduce a novel method called Gabor-based Locality Sensitive Discriminant Analysis(GLSDA), We assess the performance of our GLSDA method on face recognition task, using data sets from the ORL and Yale database. Comparative performance is carried out against other face recognition methods and the experimental results show that our method has higher recognition rate than that of other methods in the face recognition.Secondly, considering that Locality Sensitive Discriminant Analysis is a supervised dimensionality reduce algriothm, and cannot be applied when the samples’label information is unkown. Thus, here, we introduce a new method called GLLTSA, which combine the Gabor wavelet and unsupervised manifold learning method Linear Local Tangent Space Alignment. Our method can learn a linear projection which could map new data to the reduced representation space. According to the experimental results on ORL database, we also find that our method GLLTSA has higher recognition rate than that of other unsupervise manifold learning methods in the face recognition.

Related Dissertations

  1. The Classification of High Dimsnsion Flew Field Based on Manifold Learning,V231.3
  2. Research on Automatic Detection Algorithm for Substructure Distress of Highway Pavement Based on SVM,U418.6
  3. ISAR Imaging Simulation of Space Targets and Target Recognition Based on ISAR Images,TN957.52
  4. Research on Algorithms of 2D Face Template Protection,TP391.41
  5. Research on Feature Extraction and Classification of Pulse Waveform for Cholecystitis and Nephrotic Syndrome Diagnosis,TP391.41
  6. Application of Q-Learning in the Content-Based Image Retrieval Technology,TP391.41
  7. Research on Transductive Support Vector Machine and Its Application in Image Retrieval,TP391.41
  8. Research on Feature Extraction and Classification of Tongue Shape and Tooth-Marked Tongue in TCM Tongue Diagnosis,TP391.41
  9. Research on Visual Measurement for Spacecraft Rendezvous and Approach,TP391.41
  10. Research on the Image Real-Time Acquisition, Storage and Image Processing System,TP391.41
  11. Feature Extraction, Selection and Combination in Lipreading,TP391.41
  12. Multi-currency Notes Technology Research and Implementation,TP391.41
  13. The Research on Paper Currency Classification Method Based on Harr-Like Feature and Minimal Ball Including Samples,TP391.41
  14. Pavement Distress Recognition Based on Image,TP391.41
  15. Research on Visual Detection and Tracking of Mobile Robots,TP242.62
  16. Research on Fusion Algorithm of Hyper Spectral and High Spatial Resolution Remote Sensing Image,TP751
  17. An Approach for Identifying a Plant Resistance Gene Based on the Random Forest,Q943
  18. Tobacco Diseases Auto-Recognition Research Based on Image Processing Technology,S435.72
  19. Research on Nondestructive Detection Technology for External Qualities of Papayas Based-on Vision,S667.9
  20. Study of Data Reduction Technique Based on Manifold Learning,TP311.13
  21. Research on Recognition of Fabric Weave Pattern Based on Space-Frequency Domain,TS101.923

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