| 摘要: |
| 年龄是人的重要属性.近年来,自动估计用户年龄逐渐成为一个涉及模式识别、计算机视觉、语音识别、人机交互、机器学习等领域的活跃课题.其在现实世界中也有很多的实际应用,如法医学、电子商务、安全控制.日常生活中,人们往往可以很容易地根据视听信息(这里主要指人脸和语音)来判断一个人的年龄,原因在于人语音是人的年龄信息的重要载体.同样的,人机交互系统可以根据人脸图像以及语音来自动进行年龄估计.主要了基于视听信息进行年龄估计的应用领域所遇到的挑战以及现有的解决方案.详细介绍了基于视听信息的年计所用到的主要模型、算法及其性能与特点,并且分析了自动年龄估计未来可能的发展趋势. |
| 关键词: 自动年龄估计 人脸图像 说话人年龄 机器学习 |
| DOI：10.3724/SP.J.1001.2011.04012 |
| 分类号: |
| 基金项目:国家自然科学基金(60905031); 江苏省自然科学基金(BK2009269); 教育部留学回国人员科研启动基金; 模式识别国家重点实验室开放课题基金; “计算机网络与信息集成”教育部重点实验室资助项目; 东南大学优秀青年教师教学科研资助计划 |
|
| Automatic Age Estimation Based on Visual and Audio Information |
|
FANG Er-Qing, GENG Xin
|
|
School of Computer Science and Engineering, Southeast University, Nanjing 211189, China
|
| Abstract: |
| Age is an important attribute of human beings. In recent years, automatic estimations of the user’s age
have been becoming an active topic in pattern recognition, computer vision, voice recognition, human-computer
interaction (HCI), etc. It can be widely used in many real applications such as forensics, e-business, security, and so
on. In daily life, people can easily estimate the age of a person according to some visual and audio information (here
mainly refers to face and voice) because humans’ faces and voices are important agents of their age. This paper
introduces in detail the models, algorithms used in automatic age estimation based on visual and audio information,
as well as their performance and characteristics. The possible future directions for the research in automatic age
estimation are also discussed. |
| Key words: automatic age estimation face image speaker’s age machine learning |
