Face recognition based on ICA and SPSO-ELM

摘要

According to the poor robustness of the extreme learning machine as a classifier for human face recognition, a swarm optimization algorithm is proposed. The algorithm combines principle component analysis and independent component analysis to extract human face features, and extreme learning machine is used as a classifier. In order to improve the classification performance of extreme learning machine and achieve higher recognition accuracy and better robustness, the swarm optimization algorithm is introduced in the classification stage. The experimental results show that compared with the traditional algorithm, the human face recognition system using the improved algorithm not only improves the face recognition rata but also reduces the influence on the result of training data when the numbers of the hidden layer nodes change, and has good robustness, has a great promotional value in similar classification model.