Maryam Abdolali - Faculty of Computer Engineering Amirkabir University of Technology
Mohammad Rahmati - Faculty of Computer Engineering Amirkabir University of Technology

In this paper a sparse coding approach is proposed.Due to the similarity of the frequency and orientationrepresentations of Gabor filters and those of the human visualsystem, we have used Gabor filters in the step of creating thedictionary. It has been shown that not all Gabor filters in a typicalGabor bank is necessary and efficient in facial expressionrecognition. Also we proposed a voting system in the test phase ofalgorithm to find the best matching expression. The well knownJAFFE database has been used to evaluate the proposed methodand our experimental results show encouraging results within thementioned database.

Facial Expression Recognition; Sparse Coding Gabor Filters