Further insights into subspace methods with applications in face recognition
Subspace methods such as Linear Discriminant Analysis (LDA) are efficient in dimension reduction and statistical feature extraction. They are widely applied to multi-class pattern classification problems, such as face recognition, which often involve high dimensional and large data set. In this thes...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2009
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/15161 |
| _version_ | 1811684614203244544 |
|---|---|
| author | Zhu, Yan |
| author2 | Sung Eric |
| author_facet | Sung Eric Zhu, Yan |
| author_sort | Zhu, Yan |
| collection | NTU |
| description | Subspace methods such as Linear Discriminant Analysis (LDA) are efficient in dimension reduction and statistical feature extraction. They are widely applied to multi-class pattern classification problems, such as face recognition, which often involve high dimensional and large data set. In this thesis, we provide further insights into the subspace methods to resolve some prolonged issues. Firstly, we propose the Margin-Maximization Discriminant Analysis (MMDA) based on an additive-form of discriminant function, which can extract features that approximately maximize the average projected margin between the classes. Secondly, an analytical relevance measure of subspace feature vectors is derived and used to weigh the LDA features. This leads to a scheme called Relevance-Weighted Discriminant Analysis (RWDA). It completely eliminates the peaking phenomenon of LDA and also suggests a new insight into the root cause of overfitting for classifiers using distance metric. Finally, 2D subspace methods which represent images as 2D matrices are investigated, in order to tackle the computation intractability of large-scale pattern classification problems. |
| first_indexed | 2024-10-01T04:31:25Z |
| format | Thesis |
| id | ntu-10356/15161 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T04:31:25Z |
| publishDate | 2009 |
| record_format | dspace |
| spelling | ntu-10356/151612023-07-04T17:27:14Z Further insights into subspace methods with applications in face recognition Zhu, Yan Sung Eric School of Electrical and Electronic Engineering DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Biometrics Subspace methods such as Linear Discriminant Analysis (LDA) are efficient in dimension reduction and statistical feature extraction. They are widely applied to multi-class pattern classification problems, such as face recognition, which often involve high dimensional and large data set. In this thesis, we provide further insights into the subspace methods to resolve some prolonged issues. Firstly, we propose the Margin-Maximization Discriminant Analysis (MMDA) based on an additive-form of discriminant function, which can extract features that approximately maximize the average projected margin between the classes. Secondly, an analytical relevance measure of subspace feature vectors is derived and used to weigh the LDA features. This leads to a scheme called Relevance-Weighted Discriminant Analysis (RWDA). It completely eliminates the peaking phenomenon of LDA and also suggests a new insight into the root cause of overfitting for classifiers using distance metric. Finally, 2D subspace methods which represent images as 2D matrices are investigated, in order to tackle the computation intractability of large-scale pattern classification problems. DOCTOR OF PHILOSOPHY (EEE) 2009-04-08T00:56:25Z 2009-04-08T00:56:25Z 2009 2009 Thesis Zhu, Y. (2009). Further insights into subspace methods with applications in face recognition. Doctoral thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/15161 10.32657/10356/15161 en 130 p. application/pdf |
| spellingShingle | DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Biometrics Zhu, Yan Further insights into subspace methods with applications in face recognition |
| title | Further insights into subspace methods with applications in face recognition |
| title_full | Further insights into subspace methods with applications in face recognition |
| title_fullStr | Further insights into subspace methods with applications in face recognition |
| title_full_unstemmed | Further insights into subspace methods with applications in face recognition |
| title_short | Further insights into subspace methods with applications in face recognition |
| title_sort | further insights into subspace methods with applications in face recognition |
| topic | DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Biometrics |
| url | https://hdl.handle.net/10356/15161 |
| work_keys_str_mv | AT zhuyan furtherinsightsintosubspacemethodswithapplicationsinfacerecognition |