Using Facial Expression Recognition to Explore the Subject’s Reaction on Different Images
碩士 === 國立雲林科技大學 === 資訊管理系 === 106 === Face recognition is mostly applied to safety certification in current technology, but the use of facial expression recognition in life is relatively less, and the construction cost is usually relatively high. This study uses the face recognition open source tool...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2018
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/r7fvb7 |
| Summary: | 碩士 === 國立雲林科技大學 === 資訊管理系 === 106 === Face recognition is mostly applied to safety certification in current technology, but the use of facial expression recognition in life is relatively less, and the construction cost is usually relatively high. This study uses the face recognition open source tool - OpenFace, combined with the Action Units(AU) in the Facial Action Coding System as the facial expression recognition basis. Finally uses the random number method to perform the weight assignment of the expression action unit to recognize the smile, surprised and sad three expressions. The combination of features on smile and surprised has higher similarity and concentration. The key combination of sad expressions is the lip corner and the chin. In the face recognition of continuous time segment, most of the random number method can effectively judge which time zone the subject's three expressions appear.
|
|---|