| 摘要: |
| 目的 人类年龄是人类识别和搜索任务中的重要特征,现有研究一般将人脸年龄估计视为传统的分类任务,
忽略了年龄之间的有序特征,导致估计年龄与真实年龄之间的差距较大,因此,有必要寻找一种方法以缩小估计年
龄与实际年龄的差距。 方法 提出一种基于双有序性约束卷积神经网络模型(DO-CNN)的人脸图像年龄估计方法。
首先,DO-CNN 使用基于广义 Logistic 分布的有序回归模型作为卷积神经网络的分类器,并验证比其他有序分类器
在人脸估计任务上的优越性;接着,进一步提出有序竞争比损失函数,在传统竞争比损失函数上,通过引入风险项
使损失函数注意到预测年龄与真实年龄的误差,进而指导模型缩小估计年龄与真实年龄的差距。 结果 在开源人脸
图像年龄数据集 FGNET 和 AgeDB 上的对比实验显示:相比现有研究方法,DO-CNN 分别提升约 12%和 3%的准确
率,当允许的误差范围扩大后,该优势依然保持。 此外,基于广义 Logistic 分布的有序回归分类器相比基于其他分
布的有序回归分类器具有明显提升。 结论 实验结果表明:基于双有序性约束的卷积神经网络模型可以明显提升人
脸年龄估计的准确率,并减少年龄估计的实际误差。 |
| 关键词: 人脸年龄估计 有序回归 卷积神经网络 竞争比损失函数 深度学习 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| Study on Face Age Estimation Based on Double-ordinality Constraints |
|
WANG Xun, HUANG Zhensheng
|
|
School of Mathematics and Statistics, Nanjing University of Science and Technology, Nanjing 210094, China
|
| Abstract: |
| Objective Human age is an important feature in human recognition and search tasks. Existing research
generally treats age estimation in facial images as a traditional classification task ignoring the ordered characteristics of
age and resulting in a large gap between the estimated age and the actual age. Therefore it is necessary to find a method
to reduce the gap between the estimated age and the actual age. Methods This paper proposed a method for age
estimation of face images based on a double-ordinality constrained convolutional neural network DO-CNN model.
Firstly DO-CNN used an ordered regression model based on the generalized Logistic distribution as a classifier for
convolutional neural networks and verified its superiority over other ordered classifiers for face estimation tasks. Then an
ordered competitive ratio loss function was further proposed. By introducing a risk term into the traditional competitive
ratio loss function the loss function considered the error between the predicted age and the actual age thus guiding the
model to reduce the gap between the estimated age and the actual age. Results Comparative experiments on the opensource facial image age datasets FGNET and AgeDB showed that compared with existing research methods DO-CNN
improved the accuracy by about 12% and 3% respectively and this advantage remains even when allowing for a larger
error range. In addition the ordered regression classifier based on the generalized Logistic distribution exhibited significant improvements compared with ordered regression classifiers based on other distributions. Conclusion The
experimental results show that the convolutional neural network model based on double-ordinality constraints can
significantly improve the accuracy of age estimation in facial images and reduce the actual errors in age estimation. |
| Key words: face age estimation ordered regression convolutional neural networks competing ratio loss function deep
learning |
