Margin preserving projection for image set based face recognition
Face images are usually taken from different camera views with different expressions and illumination. Face recognition based on Image set is expected to achieve better performance than traditional single frame based methods, because this new framework can incorporate information about variations of...
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Conference Paper |
| Published: |
Springer
2011
|
| Online Access: | http://hdl.handle.net/20.500.11937/66262 |
| _version_ | 1848761278694162432 |
|---|---|
| author | Fan, K. Liu, Wan-Quan An, S. Chen, X. |
| author2 | na |
| author_facet | na Fan, K. Liu, Wan-Quan An, S. Chen, X. |
| author_sort | Fan, K. |
| building | Curtin Institutional Repository |
| collection | Online Access |
| description | Face images are usually taken from different camera views with different expressions and illumination. Face recognition based on Image set is expected to achieve better performance than traditional single frame based methods, because this new framework can incorporate information about variations of individual's appearance and make a decision collectively. In this paper we propose a new dimensionality reduction method for image set based face recognition. In the proposed method, we transform each image set into a convex hull and use support vector machine to compute margins between each pair sets. Then we use PCA to do dimension reduction with an aim to preserve those margins. Finally we do classification using a distance based on convex hull in low dimension feature space. Experiments with benchmark face video databases validate the proposed approach. © 2011 Springer-Verlag. |
| first_indexed | 2025-11-14T10:29:08Z |
| format | Conference Paper |
| id | curtin-20.500.11937-66262 |
| institution | Curtin University Malaysia |
| institution_category | Local University |
| last_indexed | 2025-11-14T10:29:08Z |
| publishDate | 2011 |
| publisher | Springer |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | curtin-20.500.11937-662622018-04-30T02:48:31Z Margin preserving projection for image set based face recognition Fan, K. Liu, Wan-Quan An, S. Chen, X. na Face images are usually taken from different camera views with different expressions and illumination. Face recognition based on Image set is expected to achieve better performance than traditional single frame based methods, because this new framework can incorporate information about variations of individual's appearance and make a decision collectively. In this paper we propose a new dimensionality reduction method for image set based face recognition. In the proposed method, we transform each image set into a convex hull and use support vector machine to compute margins between each pair sets. Then we use PCA to do dimension reduction with an aim to preserve those margins. Finally we do classification using a distance based on convex hull in low dimension feature space. Experiments with benchmark face video databases validate the proposed approach. © 2011 Springer-Verlag. 2011 Conference Paper http://hdl.handle.net/20.500.11937/66262 10.1007/978-3-642-24958-7_79 Springer restricted |
| spellingShingle | Fan, K. Liu, Wan-Quan An, S. Chen, X. Margin preserving projection for image set based face recognition |
| title | Margin preserving projection for image set based face recognition |
| title_full | Margin preserving projection for image set based face recognition |
| title_fullStr | Margin preserving projection for image set based face recognition |
| title_full_unstemmed | Margin preserving projection for image set based face recognition |
| title_short | Margin preserving projection for image set based face recognition |
| title_sort | margin preserving projection for image set based face recognition |
| url | http://hdl.handle.net/20.500.11937/66262 |