Human Facial Age Estimation Via Deep Forest And Hierarchical Regression

Human facial age estimation has received continuous attention due to its wide real-world applications in security control,age-targeted advertising,biometrics,etc.In this paper,a novel deep forest approach termed fcXGBoost is proposed for the regression task of human facial age estimation.The fcXGBoost approach consists of Cascade XGBoost with a cascade structure for representation learning,and multi-feature fusion for the enhancement of its representational learning ability with fusion of multiple features,including Active Appearance Models(AAM),Local Binary Patterns(LBP)and Gabor Wavelets(GW).A hierarchical regression approach is proposed to further improve the estimation performance,which consists of two stages:a rough regression stage and a detailed regression stage.In the rough regression stage,we classify a sample into a determined ages group according to the value of the predicted age by a fcXGBoost.While in the detailed regression stage,we estimate the accurate age by a special fcXGBoost.Moreover,a flexible overlapping of age ranges in the detailed age estimation is proposed to eliminate the influence of error in rough regression stage.The proposed approach is evaluated on the benchmark and public facial image dataset of both MORPH Album 2 and its subset by comparison to the state-of-the-art results.