Age is a human attribute which grows alongside an individual. Estimating human age is quite difficult for machine as well as humans, however there has been and are still ongoing efforts towards machine estimation of human age to a high level of accuracy. In a bid to improve the accuracy of age estimation from facial image, several approaches have been proposed many of which used Machine Learning algorithms. The several Machine Learning algorithms employed in these works have made significant impact on the results and of performances of the proposed age estimation approaches. In this paper, we examined and compared the performance of a number of Machine Learning algorithms used for age estimation in several previous works. Considering two publicly available facial ageing datasets (FG-NET and MORPH) which have been mostly used in previous works, we observed that Support Vector Machine (SVM) has been most popularly used and a combination/hybridization of SVM for classification (SVC) and regression (SVR) have shown the best performance so far. We also observed that the face modelling or feature extraction techniques employed significantly impacted the performance of age estimation algorithms.

The task of estimating the age of humans from facial image is a challenging one due to the non-linear and personalized pattern of aging differing from one individual to another. In this work, we investigated the problem of estimating the age of humans from their facial image using a GroupWise age ranking approach complemented by ageing pattern correlation learning. In our proposed GroupWise age-ranking approach, we constructed a reference image set grouped according to ages for each individual in the reference set and used this to obtain age-ranks for each age group in the reference set. The constructed reference set was used to obtain transformed LBP features called age-rank-biased LBP (arLBP) features which were used with attached age-ranks to train an age estimating function for predicting the ages of test images. Our experiments on the publicly available FG-NET dataset and a locally collected dataset (FAGE) shows the best known age estimation accuracy with MAE of 2.34 years on FG-NET using the leave-one-person-out strategy.