Feature Extraction Techniques For Facial Micro-Expression Recognition
Feature extraction techniques play a significant role in many computer vision tasks such as detection and recognition. To be able to effectively describe targets, a suitable feature extraction method has to be applied. In this research work, the main goal is to design or formulate the feature extract...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 出版: |
2016
|
| 主題: | |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | Feature extraction techniques play a significant role in many computer vision tasks such as detection and recognition. To be able to effectively describe targets, a suitable feature extraction method has to be applied. In this research work, the main goal is to design or formulate the feature extraction techniques for facial micro-expression recognition. Subtle emotions possess distinct characteristics compared to the normal facial expressions in a few aspects: elapsed duration and motion intensity. For facial micro-expressions, they are subtle (i.e. less intensive or obvious facial motion changes) and short elapsed duration. Thus, to capture subtle emotions, the designed features have to be able to: (1) contain both spatial and temporal information and (2) preserve the locality information (as facial micro-expressions usually occur at one part of a face). This dissertation introduces three proposed spatio-temporal feature extraction techniques for facial micro-expression recognition based on videos. |
|---|