Spatio-temporal framework and algorithms for video-based face recognition
The objective of this research is to devise algorithms for video-based face recognition based on spatio-temporal manifolds. The primary focus is to better represent nonlinear face patterns on a data manifold, while incorporating spatio-temporal information that is inherent in videos. In addition, a...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 出版: |
2014
|
| 主題: | |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | The objective of this research is to devise algorithms for video-based face recognition based on spatio-temporal manifolds. The primary focus is to better represent nonlinear face patterns on a data manifold, while incorporating spatio-temporal information that is inherent in videos. In addition, a systematic approach to evaluate sampling is also crucial as current experimental setups are inadequate to provide balanced evaluation. These goals are inspired by motivations from computer vision, psychology and cognitive neuroscience. |
|---|