Gender Classification Based on Evolutionary Extreme Learning Machine

We introduce an evolutionary extreme learning machine (E-ELM) and incremental bilateral two-dimensional principal component analysis (IB2DPCA) for gender classification, a novel algorithm for face feature extraction, which directly extracts the proper features from image matrices. There are both ways of contributions in this paper. First, we proposed the IB2DPCA-EELM, the IB2DPCA being our new idea. Second, we showed that it can be another feasible tool for E-ELM of gender classification. The proposed method is based on curvelet image decomposition of human faces and a subband that exhibits a maximum standard deviation is dimensionally reduced using an improved dimensionality reduction technique. Using IB2DPCA learns from a small pool of labeled data and then iteratively selects the most informative samples from the unlabeled set to increasingly improve the classifier accuracy. Other notable contributions of the proposed algorithm include significant improvements in gender classifier, extreme reduction in training time and minimal dependence on the number of prototypes. Extensive experiments are performed using challenging databases and results produce better performance compared with other algorithms.