Improved softmax loss for deep learning-based face and expression recognition

Improved softmax loss for deep learning-based face and expression recognition

In recent years, deep convolutional neural networks (CNN) have been widely used in computer vision and significantly improved the performance of image recognition tasks. Most works use softmax loss to supervise the training of CNN and then adopt the output of last layer as features. However, the dis...

Full description

Bibliographic Details
Main Authors: Jiancan Zhou, Xi Jia, Linlin Shen, Zhenkun Wen, Zhong Ming
Format: Article
Language:English
Published: Wiley 2019-09-01
Series:Cognitive Computation and Systems
Subjects:
gradient methods
image recognition
computer vision
face recognition
learning (artificial intelligence)
feature extraction
convolutional neural nets
deep convolutional neural networks
improved softmax loss
deep cnn
expression recognition
deep learning-based face recognition
mnist handwritten digits dataset
casia-webface
fer2013 training set
lfw dataset
discriminative features
Online Access:https://digital-library.theiet.org/content/journals/10.1049/ccs.2019.0010

Internet

https://digital-library.theiet.org/content/journals/10.1049/ccs.2019.0010

Similar Items