Abstract
Recently, vast attention has grown in the field of computer vision, especially in face recognition, detection, and facial landmarks localization. Many significant features can be directly derived from the human face, such as age, gender, and race. Estimating the age can be defined as the automatic process of classifying the facial image into the exact age or to a specific age range. Practically, age estimation from the face is still a challenging problem due to the effects from many internal factors, such as gender and race, and external factors, such as environments and lifestyle. Huge efforts have been addressed to reach an accepted and satisfied accuracy of age estimation task. In this paper, we try to analyze the main aspects that can increase the performance of the age estimation system, present the handcrafted-based models and deep learning-based models, and show how the evaluations are being conducted, discuss the proposed algorithms and models in the age estimation, and show the main limitations and challenges facing the age estimation process. Also, different aging databases that contain age annotations are discussed. Finally, few guidelines and the future prospect related to the age estimation are investigated.
Highlights
Human age is considered as a significant personal feature that directly can be derived from the emerging of different patterns of the facial appearance [1]
This review aims to: 1- Outline all the issues related to Age Estimation (AE) system. 2- Compare between the traditional methods with the most recent technologies which used deep learning
The most common and available datasets that have age annotations have been offered in this review
Summary
Human age is considered as a significant personal feature that directly can be derived from the emerging of different patterns of the facial appearance [1]. The problem of estimating the facial age has the same challenges as other facial image recognition tasks While they need firstly to detect the face to locate the main facial features related to the task, later the feature vector should be formulated and the image will be classified [2]. Different restrictions related to the age are needed to be applied on a physical access or virtual entrée to a website or mobile application. Age estimation models can be based on handcrafted algorithms or deep learning technology. An intermediate case between regression and classification is called soft classification [14] which based on the gaussian distribution centered at the target age, was used to estimate the age from facial image.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
