Paper Title
Comparison Study on Facial Expression Recognition
Abstract
This letter considers the issue of automatic facial expressions recognition. In this work, we challenge the given
benchmark protocol of the Extended Cohn-Kanade (CK+) database, and we used Gabor filters with the Sparse
Representation based Classifier (SRC) to outperform their best result by 5 percentage points; we propose a new cropping
technique for emotional faces, and we consider different combination of appearance-based projection methods, feature
extraction techniques and classifiers to make a fair comparison among them. Experimental results underline the superior
performance of SRC, particularly successful when coupled with Gabor features
Keywords - Facial Expression Recognition, Sparse Representation based Classifier, Features Extraction, Gabor Filters.