A study on predicting age using Convolution Neural Network CNN

Automatic age segregation has become important in many programs, especially since the proliferation of social media and social media platforms. By reading presentations through deep-convolution neural networks (CNN) networks, significant improvements in performance can be seen in these activities. A quick and effective system for measuring gender and age based on facial expressions is developed. Following multiple tests using public face data, our proposed method proved to be have shown way better results than current methods available in literature. Training phase of our proposed model changes the assemble very rapidly of the network we use, therefore our model had achieved greater accuracy while reckoning the age using facial data in years. Our model also consumes the minimum network model area. Our work can be summarized in four main points: (1) Development of a lightweight convolution network based on mixed mixing process to find age. (2) Age ratio accuracy is high, that leads to an appropriate method of division combining and conversion is easy. (3) Image magnification is use to facilitates the face detection and adjustment of the image input, this leads to enhancement of age related features.