Independent Component Regularisation Analysis on Face Verification
Face verification system is one of the biometrics technologies which has gained considerable attention since the past decades. It uses human’s unique characteristic which is face to authenticate one’s identity. Contactless verification from users to the device and avoiding using expensive devices ar...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 出版: |
2015
|
| 主題: | |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | Face verification system is one of the biometrics technologies which has gained considerable attention since the past decades. It uses human’s unique characteristic which is face to authenticate one’s identity. Contactless verification from users to the device and avoiding using expensive devices are two main reasons for the popularity of face verification application. Illumination, pose and facial expressions variations always resulting in performance degradation. Resolving these intra-class variations for promising performance is still a challenge in nowadays. Dimensionality reduction is one of the main tasks in face verification. Instead of using high dimensionality facial images, subspace approach has been introduced whereby each image dimension is reduced while redundant features are eliminated. This approach attempts to seek the most appropriate representation of the face data and extract features without sacrificing its discriminant information. |
|---|