Research on Facial Age Prediction Based on Feature Selection and Manifold Learning
In recent years,due to the widespread application of human appearance age estimation in personal safety and law enforcement,it has become a hot research topic and has attracted widespread attention from academia and the market.In order to achieve the goal of age estimation,a large number of scholars have proposed relevant methods.Among them,models derived through cumulative attribute encoding achieve good performance by preserving the proximity similarity of age.However,the above methods ignore the geometric structure of the extracted facial features.In fact,the geometric structure of data greatly affects the accuracy of predictions.Therefore,this article proposes an age estimation algorithm called feature selection and geometric preserving least squares regression,which combines feature selection and manifold learning paradigms.Compared with other methods,our proposed method not only preserves the geometric structure in facial representation,but also selects distinctive features.Finally,relevant experiments were conducted on the Morph2 dataset,and the results proved that the method achieved optimal performance.
facial age predictionaccumulated attribute encodingfeature selectionmanifold learning
NETL
NSTL
万方数据
