Facial Age Estimation Using a Multi-task Network Combining Classification and Regression

Abstract

Age estimation of facial images is very challenging because of the complexity of face aging process and the difficulty of collecting and labeling data. A holistic regression model is subject to imbalanced training data, while a divide-and-conquer method highly depends on the effect of the age classification, which usually has boundary effect due to cross-age correlations. This paper proposes a simple but effective multi-task learning (MTL) network combining classification and regression for age estimation called CR-MT net, where classification acts as an auxiliary task to regression. MTL can boost the generalization performance of the age regression task by shared information representation learning of the two tasks. Compared to divide-and-conquer methods, our method performs MTL for two tasks with an end-to-end training, and has no error propagation from classification to regression. The holistic regression model in CR-MT net does not meet the boundary effect, and can fit the heterogeneous or unbalanced age data more accurately with the aid of a good age data partition in classification. We evaluate two age grouping techniques to find a good data partition, and diagnose various factors which can influence the performance of the CR-MT net by extensive experiments. The CR-MT net is verified in three public datasets, and achieves state-of-art results. INDEX TERMS Age estimation, multi-task learning, deep learning.

Organization, Extraction, Classification and Prediction of Age in Facial Images using Convolutional Neuronal Network

“…The period of 44 -57 years of age, obtained an accuracy of 0.90 and a low Recall, thus indicating that such a period is not adequately detected by the mathematical model, being the period that lower Recall has, on the other hand, the period of 15 -23 years of age, has a good accuracy of 0.95 and a Recall of 0.95 thus indicating that the algorithm detects it well. There are multiple research articles, which have performed facial image analyses such as [18], [19], [20], [21] and [2] in these works as well as in many others the images are analyzed, with different methods such as CNN, ordinal method of deep learning, directional age patterns, Kullback-Leibler divergence and antagonistic generative networks; all these methods show very good results, as well as the one obtained in between work; in the reference [4] they obtain an average success rate of approximately 88%, with a CNN model, as well as in the reference [22], where using a machine learning structure they manage to reduce by up to 37.75% the error finally estimated in [23] use large amounts of data to achieve results superior to those shown in this work. the results of our research show that we are on the right track since we have obtained an efficiency in the classification greater than 97.86%, and an efficiency in predictions approximately 87.64%, these values indicate that we are within an acceptable range of efficiency in the classification, however, it is necessary to improve even more, therefore, this work will continue to be developed to obtain greater efficiency in predictions.…”

Multi‐task multi‐scale attention learning‐based facial age estimation

“…Liu et al. [8] developed a simple but effective multi‐task learning (MTL) network, which boosted the generalisation performance of the age regression task by shared information representation learning of the two tasks. Han et al.…”

Deep Learning Based Real Age and Gender Estimation from Unconstrained Face Image towards Smart Store Customer Relationship Management

“…A lightweight CNN network with mixed attention mechanism for low end devices was proposed in [54], where the output layer was fused by classification and regression approach. Another multi-task learning approach merging classification and regression concepts to fit the age regression model with heterogeneous data with the help of two different techniques for partitioning data towards classification was proposed in [55]. To resolve the problem of data disparity and ensure the generality of the model, a very recent method is proposed by Kim et al [56] where a cycle generative adversarial network-based race and age image transformation method is used to generate sufficient data for each distribution.…”