Facial Expression Classification using Multi-Scale Histogram of Oriented Gradients
Abstract
An automatic facial expression classification method, based on multiscale Histogram of Oriented Gradients (HOG) features extracted from sub-facial image patches, is proposed in this paper. A number of multiclass Support Vector Machines (SVM) are designed using the multiscale HOG features for classifying six basic facial expressions for frontal face images. The proposed method, with a simple design and few training data, achieves descent classification accuracy at a very low computational complexity.
Keywords: multi-scale HOG, Support Vector Machine, Sub-facial patch, histogram.
Cite this Article: Sagar Deep Deb, Manish Sharma, Chandrajit Choudhury, Fazal Ahmed Talukdar, Rabul Hussain Laskar. Facial Expression Classification using Multi-Scale Histogram of Oriented Gradients. International Journal of Image Processing and Pattern Recognition: 2020; 6(1): 5–13p.
Full Text:
PDFDOI: https://doi.org/10.37628/ijoippr.v6i1.564
Refbacks
- There are currently no refbacks.