JIITA, vol. 2, no. 4, pp.203-207, Dec, 2018
Age and Gender Estimation based on Multiple patterns and Multiple features
Hyunduk Kim*,1), Sang-Heon Lee1), Byunghun Hwang1), Yoon-Jib Kim1), Youngsun Ahn1)
1) Daegu Gyeongbuk Institute of Science & Technology
Abstract: Human age and gender are valuable demographic characteristics. They are also important soft biometric traits useful for human identification or verification. In this paper, we propose an age and gender estimation framework based on multiple patterns and multiple features. The proposed approach consists of three process. First step is landmark based face alignment. Second step is feature extraction step. In this process, we define multiple patterns from face region. And then, we combine multiple features extracted from multiple patterns. Finally, we classify age and gender using a multi-layered Support Vector Machines (SVM) for efficient classification. Rather than performing gender estimation and age estimation independently, the use of the multi-layered SVM can improve the classification rate by constructing a classifier that estimate the age according to gender. Moreover, we collect a dataset of face images, called by DGIST_C, from the internet. Our dataset consists of about 20K Korean celebrities, labeled for age, gender.
Keyword: age estimation; gender estimation; feature fusion approach; multi-layered approach