Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Neighborhood Loss for Age Estimation from Face Image Using Convolutional Neural Networks


Affiliations
1 Institute of Information Technology, High-Tech Research and Development Centre, Kim Il Sung University, Korea, Democratic People's Republic of
     

   Subscribe/Renew Journal


Convolutional Neural Network (CNN) is widely used in estimating age from face image. In many CNN applications such as image classification, face recognition and other computer vision scopes, the cross-entropy loss is used as a supervision signal to train CNN model. However, the cross-entropy loss only enhances the separability of classes and does not consider their correlation in age estimation task. In this paper we propose a novel loss function called neighborhood loss which regards the correlation between classes in age estimation by modifying standard cross entropy loss. To evaluate the effectiveness of the proposed neighborhood loss, we present CNN architecture based on the residual units. Through some experiments, we show that neighborhood loss provides superior performance compared to prior works in age estimation.

Keywords

Age Estimation, Neighborhood Loss, Convolutional Neural Network.
Subscription Login to verify subscription
User
Notifications
Font Size


  • Neighborhood Loss for Age Estimation from Face Image Using Convolutional Neural Networks

Abstract Views: 250  |  PDF Views: 1

Authors

Hyok Kwak
Institute of Information Technology, High-Tech Research and Development Centre, Kim Il Sung University, Korea, Democratic People's Republic of
Chol Nam Om
Institute of Information Technology, High-Tech Research and Development Centre, Kim Il Sung University, Korea, Democratic People's Republic of
Il Han
Institute of Information Technology, High-Tech Research and Development Centre, Kim Il Sung University, Korea, Democratic People's Republic of
Jang Su Kim
Institute of Information Technology, High-Tech Research and Development Centre, Kim Il Sung University, Korea, Democratic People's Republic of

Abstract


Convolutional Neural Network (CNN) is widely used in estimating age from face image. In many CNN applications such as image classification, face recognition and other computer vision scopes, the cross-entropy loss is used as a supervision signal to train CNN model. However, the cross-entropy loss only enhances the separability of classes and does not consider their correlation in age estimation task. In this paper we propose a novel loss function called neighborhood loss which regards the correlation between classes in age estimation by modifying standard cross entropy loss. To evaluate the effectiveness of the proposed neighborhood loss, we present CNN architecture based on the residual units. Through some experiments, we show that neighborhood loss provides superior performance compared to prior works in age estimation.

Keywords


Age Estimation, Neighborhood Loss, Convolutional Neural Network.

References