Three Level Optimization Models of Scaled Gabor Features for Facial Expression Recognition
Neha, Pratistha Mathur · International Journal of Engineering & Technology · 2018
The area of computer vision and machine learning for pattern recognition has witnessed the need for research for the development of algorithms for different applications such as human-computer interaction, automated access control and surveillance. In the field of computer vision Facial Expression Recognition has attracted the researcher’s interest. This paper presents a novel feature extraction technique: Gabor-Average-DWT-DCT for automatic facial expression recognition from a person's face image invariant of illumination. Facial Emotions have different edge and texture pattern. Gabor filter is able to extract edges and texture pattern of faces but with problem of huge dimension and high redundancy. The problem of huge dimension and high redundancy is reduced by proposed Average-DWT-DCT feature reduction technique in order to increase accuracy of system. Proposed Gabor- Average -DWT-DCT provides a compact feature vector for reducing response time of system compared to existing Gabor based expression classification. Detailed quantitative analysis is done and results that the average recognition rate of proposed technique is better than state of art results.