Cascade support vector regression-based facial expression-aware face frontalization

Yiming Wang, Hui Yu, Honghai Liu, Junyu Dong, Muwei Jian

Research output: Chapter in Book/Report/Conference proceedingConference contribution

312 Downloads (Pure)


The main aim of face frontalization is to synthesize the frontal facial appearances from non-frontal facial images. How to estimate the frontal face-shape is a crucial but very challenging problem in the frontalization task. Most existing methods use a single frontal face-template to fit in with frontal facial appearances, which will result in a loss of expression related information. In this work, we present a novel facial expression-aware face frontalization method which directly learns the pair-wise relations between non-frontal face-shape and its frontal counterpart. The support vector regression is explored to train the pair-wise relation model. Considering non-lineariality of the relationship, an appropriate cascade manner is applied to iteratively adjust and optimize the model. The frontal face-shape is then estimated via this model. With the estimated shape, frontal appearances are synthesized through a texture-fitting process formulated by solving a simple optimization problem. The proposed method has been evaluated on a in-the-wild facial expression database. The experimental results shows an outstanding performance of both facial expression-aware frontal face recovery and facial expression recognition.
Original languageEnglish
Title of host publication2017 IEEE International Conference on Image Processing (ICIP)
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)978-1509021758
ISBN (Print)978-1509021765
Publication statusPublished - 22 Feb 2018
Event24th IEEE International Conference on Image Processing - Beijing, China
Duration: 17 Sept 201720 Sept 2017

Publication series

NameIEEE ICIP Proceedings Series
ISSN (Electronic)2381-8549


Conference24th IEEE International Conference on Image Processing


  • face frontalization
  • facial expressionaware
  • facial expression recognition
  • support vector regression
  • facial expression analysis
  • RCUK
  • EP/N025849/1


Dive into the research topics of 'Cascade support vector regression-based facial expression-aware face frontalization'. Together they form a unique fingerprint.

Cite this