Title: A New Feature Descriptor for Face Recognition

Year of Publication: 2015
Page Numbers: 83-85
Authors: Gülden Eleyan, Alaa Eleyan
Conference Name: The International Technology Management Conference (ITMC2015)
- Turkey


The idea behind this paper is to use a new statistical feature for the face recognition problem. To the best of our knowledge, no one has attempted to implement this approach before. The idea is simple and straight forward. For each face image, a feature vector is formed by concatenating 4 vectors together namely; row-wise sum, column-wise sum, diagonalwise sum and antidiagonal-wise sum. The generated feature vector is then used directly for classification or is can be first projected to another space such as eigenspace or fisherspace. The preliminary simulation results on ORL database show that the proposed approach outperforms the well-known principal component analysis (PCA) algorithm for face recognition.