基于改进深度残差网络的人脸年龄估计
首发时间:2020-12-18
摘要:目前基于深度学习的人脸年龄估计成为人脸研究领域的一个热点,由于导致人脸衰老的不确定性因素众多且复杂,同时待估计的人脸图像容易受噪声影响。本文将ResNet与上下文特征模组相结合并提出基于改进的ResNet的人脸年龄估计网络,并在训练时对训练图像进行模拟真实环境背景的处理,经过在UTK Face和LAP 2015数据集上的实验证明,本文提出的网络可以较好地实现人脸年龄估计。
For information in English, please click here
Face age estimation based on improved deep residual network
Abstract:At present, face age estimation based on deep learning has become a hot spot in the field of face research. Because the uncertain factors that lead to face aging are numerous and complex, and the face image to be estimated is easily affected by noise.In this paper, ResNet is combined with context feature module, and a face age estimation network based on improved ResNet is proposed, and the image is processed by simulating the real environment background during training. Experiments on UTK Face and LAP 2015 data sets show that the network proposed in this paper can better realize face age estimation.
Keywords: Face age estimation Deep residual network Deep learning Context module
引用
No.****
动态公开评议
共计0人参与
勘误表
基于改进深度残差网络的人脸年龄估计
评论
全部评论0/1000