Face age estimation method and system based on sparse undirected probabilistic graphical model

Abstract

The invention discloses a face age estimation method based on a sparse undirected probabilistic graphical model and belongs to the technical fields of machine learning and mode identification.According to the face age estimation method, an age distribution prediction model is innovatively constructed through an undirected probabilistic graph, and in addition, proper sparsity regular terms are added to an optimization training target of the model so as to constrain model parameters.The invention further discloses a face age estimation system based on the sparse undirected probabilistic graphical model.Compared with the prior art, the face age estimation method and system have the advantages that more abundant information can be learned out of complex image features so as to predict age distribution, and in addition, more compact encoding can be conducted on the information through term vectors; the experience of image sparsity is utilized, and the sparsity regular terms are added to constrain the model parameters, so that model obtained through learning has better generalization.

Description

technical field [0001] The invention relates to a face age estimation method, in particular to a face age estimation method based on a sparse undirected probability graph model, which belongs to the technical field of machine learning and pattern recognition. Background technique [0002] The application of automatic age estimation based on face images is becoming more and more widely, mainly including the following aspects: (1) Age-based human-computer interaction system: on the basis of ordinary human-computer interaction system, the human age automatic estimation algorithm is introduced, according to the user's (2) age-based access control system: used to prevent minors from accessing inappropriate web pages or content, purchasing tobacco and alcohol products on vending machines, entering bars, etc. Suitable places, etc.; (3) E-commerce: Estimate the approximate age of customers based on images, etc., and adopt different marketing strategies for customers of different age...

AI Technical Summary

Problems solved by technology

One is that for image data, the extracted image features are often very complex, and BFGS-LLD is based on the maximum entropy model, and the maximum entropy model is a relatively simple parameter model, which cannot learn enough information to predict age distribution

The second problem is that this model does not use the prior of image sparsity. In many previous studies and practices, sparsity has been proved to be a very useful prior for image features, but the previously proposed age marker The distribution model cannot take advantage of such effective prior knowledge, so the learned prediction model lacks convincingness and credibility

Images