Age Estimation Based on Facial Images Using Hybrid Features and Particle Swarm Optimization

Face images provide significant biological information. Face age estimation is one of the primary research directions in face images. The appearance of the face changes dynamically, and these changes are influenced by a variety of factors such as light, aging, makeup, beard, etc. The human face has many characteristics, including emotions, sex, race, age, etc. Face age estimation has numerous applications in biometrics, security, commercial, military, interaction with computers, and providing services to the individual. In this paper, the age estimation system is divided into four distinct stages. The first step is to extract local features including Gabor wavelets (GW), Local Binary Pattern (LBP), Local Phase Quantization (LPQ), and Histogram of Oriented Gradients (HOG). These attributes are then combined in the second step as a feature fusion method, which combines four different feature extraction methods. In the following step, Particle Swarm Optimization (PSO) as a meta-heuristic optimization algorithm is used to decrease the size of attributes and find optimal features. Finally, we used classification and regression methods to estimate the age and age groups. Initially, we used support vector machines (SVMs) to determine the age class, and then used support vector regression (SVRs) to estimate the ages within those groups. Finally, our algorithm was examined on two widely used databases, i.e. the FG-NET and the MORPH, to determine its effectiveness regarding aging estimates. In the FGNET dataset, we achieved an MAE of 3.34 years and 75.69 percent classification accuracy, and in the MORPH dataset, we obtained an MAE of 3.21 years and 81.66 percent classification accuracy.