A 2D Personalized Facial Expression Generation Approach

  • Authors:
  • Xinjuan Zhu;Xue Li;Lianjie Song

  • Affiliations:
  • Department of Computer Science, Xi'an Polytechnic University, China;Department of Computer Science, Xi'an Polytechnic University, China;Department of Computer Science, Xi'an Polytechnic University, China

  • Venue:
  • Proceedings of the Second International Conference on Innovative Computing and Cloud Computing
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper we propose an approach for generating personalized facial expression image from one expressionless full-face photo. For each given expressionless photo, image standardization needs to be done. The feature points of facial expression are defined and extracted. And neural network model is built to predict the facial feature point relative offset of expression image. Then the prediction results are used to process the deformation from expressionless image to the expression image. The happy and angry expression binary images are automatically generated from expressionless binary images in our experiment. Experimental results show that the proposed method can generate realistic and personalized facial expression images in real time. The work will offers a reference to the future research for personalized 2D web virtual human facial expression animation generation.