A Novel Sparse Representation-based Face Recognition System Using Salient Feature
Aihua Yu, Gang Li, Beiping Hou, Hongan Wang, Gaoya Zhou · 2018
Sparsifying dictionary learning aims at finding a dictionary in which the training data admits a sparse representation. Such a representation technique has been used in face recognition. The conventional patch based dictionary learning framework is not appropriate for the application of on-line face recognition due to the huge amount of data involved. A novel sparse representation-based face recognition scheme using salient feature is proposed in this paper. As a weighting strategy is adopted, such a scheme is not only more efficient but also more robust to the misalignment and image noises and hence improves recognition performance. Experiments on ORL and FERET databases validate the proposed face recognition system.