Microexpression recognition based on improved robust principal component analysis and texture feature extraction

Dong Xiaochen, Zhao Zhigang, Li Qiang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Micro-expression can reflect the emotional information that humans can hardly conceal. There are three main characteristics: short duration, low intensity and local movement. From these characteristics it can be seen that the motion of the microexpression is sparse. For sparse micro-expression movement, a robust principal component analysis (RPCA) was proposed to extract subtle micro-expression motion information. Using improved Edge Direction Histogram (EOH) algorithm and Binary Gradient Contours (BGC) algorithm to extract local texture features can solve the problem of spatio-temporal domain and obtain high recognition accuracy. Experiments on the SMIC database show that the proposed algorithm has better performance.

Original languageEnglish
Title of host publicationICCIP 2018 - Proceedings of 2018 4th International Conference on Communication and Information Processing
PublisherAssociation for Computing Machinery
Pages48-53
Number of pages6
ISBN (Electronic)9781450365345
DOIs
StatePublished - 2 Nov 2018
Externally publishedYes
Event4th International Conference on Communication and Information Processing, ICCIP 2018 - Qingdao, China
Duration: 2 Nov 20184 Nov 2018

Publication series

NameACM International Conference Proceeding Series

Conference

Conference4th International Conference on Communication and Information Processing, ICCIP 2018
Country/TerritoryChina
CityQingdao
Period2/11/184/11/18

Keywords

  • BGC algorithm
  • EOH algorithm
  • Recognition accuracy
  • Robust principal component analysis

Fingerprint

Dive into the research topics of 'Microexpression recognition based on improved robust principal component analysis and texture feature extraction'. Together they form a unique fingerprint.

Cite this