A new age estimation method based on ensemble learning

For the last decade, research on face recognition has attracted much attention due to its wide application. How-ever, the factor that affects the performance of face recognition and has been neglected by researchers is the aging variation. Age estimation is an important problem in the face recognition technique. This paper proposes two algorithms for age estimation by utilizing canonical correlation analysis and cost-sensitive learning. Furthermore, these two algorithms are combined into the ensemble method to improve the accuracy of age estimation. Finally, the experimental results show the effectiveness of the proposed method.