Abstract
Facial Expressions are one of the main methods we use to express our emotions to others. Yet Facial Expression Recognition (FER) remains a difficult topic for machines to intrepret. While Computer Vision can extract features quite easily from imagery, there is still the difficult step of recognizing what emotion those features belong to. Many have taken to Deep Learning to bridge this learning gap. However this paper shows that with selected features, even classic techniques without modification can achieve high accuracy. This paper demonstrates how select features, taken from ANOVA, LDA and PCA, enhances the accuracy of HOG without further processes.
Original language | English |
---|---|
Number of pages | 8 |
Publication status | Accepted/In press - 6 Jul 2018 |
Event | Irish Machine Vision and Image Processing Conference - Belfast, United Kingdom Duration: 29 Aug 2018 → 31 Aug 2018 |
Conference
Conference | Irish Machine Vision and Image Processing Conference |
---|---|
Abbreviated title | IMVIP |
Country/Territory | United Kingdom |
Period | 29/08/18 → 31/08/18 |
Keywords
- Facial Expression Recognition
- Feature Selection
- Feature Reduction
- Machine Vision
- Machine learning