【摘要】 年龄是人脸的重要生物属性之一,在人机交互、安全监控等很多领域具有广泛的应用价值,但是由于人脸包含身份、年龄、性别、种族、表情、姿态等众多相关属性,并且每个人的年龄演化进程各不相同,因此年龄估计仍是一个非常具有挑战性的科学课题.文中从人脸图像年龄估计研究的问题定义、数据集采集、特征提取、模型表示等角度,综述当前年龄估计的主要研究工作和进展,并进一步探讨年龄估计研究主要挑战以及未来潜在的研究方向.更多还原

【Abstract】 As one of the most important biometric attributes of human face, age information has comprehensive application value in many areas such as human computer interaction, safety surveillance and so on. However, there are many related attributes on the face including identity, age, gender, race, expression, pose and so on, and people usually have different ways of aging process. Therefore, age estimation is still a great challenge. This paper investigates the existing state-of-the-art age estimation methods from the viewpoint of the definition of facial image based age estimation problem, dataset collection, feature extraction and model representation. Finally, the challenges and potential directions of age estimation are discussed.更多还原