FLD-SIFT: Class Based Scale Invariant Feature Transform for Accurate Classification of Faces
In this paper, a new model called FLD-SIFT is devised for compact representation and accurate recognition of faces. Unlike scale invariant feature transform model that uses smoothed weighted histogram and massive dimension of feature vectors, in the proposed model, an image patch centered around the keypoint has been considered and linear discriminant analysis (FLD) is employed for compact representation of image patches. Contrasting to PCA-SIFT model that employs principal component analysis (PCA) on a normalized gradient patch, we employ FLD on an image patch exists around the keypoints. The proposed model has better computing performance in terms of recognition time than the basic SIFT model. To establish the superiority of the proposed model, we have experimentally compared the performance of our new algorithm with (2D)2-PCA, (2D)2-FLD and basic SIFT model on the AT&T face database.
KeywordsLinear discriminant analysis Local descriptor Face classification
Unable to display preview. Download preview PDF.
- 2.Harris, C., Stephens, M.: A combined corner and edge detector. In: Alvey Vision Conference, pp. 147–151 (1988)Google Scholar
- 10.Van Gool, L., Moons, T., Ungureanu, D.: Affine/photometric invariants for planar intensity patterns. In: Buxton, B.F., Cipolla, R. (eds.) ECCV 1996. LNCS, vol. 1065. Springer, Heidelberg (1996)Google Scholar
- 14.Mikolajczyk, K., Schmid, C.: A performance evaluation of local descriptors. In: Proceedings of Computer Vision and Pattern Recognition (June 2003)Google Scholar
- 15.Mikolajczyk, K., Schmid, C.: Indexing based on scale invariant interest points. In: Proceedings of International Conference on Computer Vision, pp. 525–531 (July 2001)Google Scholar
- 19.Ke, Y., Sukthankar, R.: PCA-SIFT: A more distinctive representation for local image descriptors. In: Computer Vision and Pattern Recognition (2004)Google Scholar