Presentation 2014-10-17
Unsupervised learning from facial expression by Latent Dirichlet Allocation:Discover the hidden topics from a face
Prarinya SIRITANAWAN, Tu Bao HO, Kazunori KOTANI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Given a set of unlabeled data, it is desired to retrieve the significant information hidden in the messy data. In this research, we utilize the Latent Dirichlet Allocation algorithm to discover the hidden topic from the set of unlabeled facial expression images. Traditionally, the learning of facial expression is supervised by a well-labeled dataset. The features are separated into six emotions according to the psychological research. However, there is a lot of hidden information in our facial expression which does not follow those rules. Therefore, the question arises of how many possible facial expressions human can express. In this research, we retrieve the latent topics to explore the characteristic of the complex facial expression.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) topic modelling / facial expression / Latent Dirichlet Allocation / clustering
Paper # BioX2014-37
Date of Issue

Conference Information
Committee BioX
Conference Date 2014/10/9(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Biometrics (BioX)
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Unsupervised learning from facial expression by Latent Dirichlet Allocation:Discover the hidden topics from a face
Sub Title (in English)
Keyword(1) topic modelling
Keyword(2) facial expression
Keyword(3) Latent Dirichlet Allocation
Keyword(4) clustering
1st Author's Name Prarinya SIRITANAWAN
1st Author's Affiliation Japan Advanced Institute of Science and Technology()
2nd Author's Name Tu Bao HO
2nd Author's Affiliation Japan Advanced Institute of Science and Technology
3rd Author's Name Kazunori KOTANI
3rd Author's Affiliation Japan Advanced Institute of Science and Technology
Date 2014-10-17
Paper # BioX2014-37
Volume (vol) vol.114
Number (no) 251
Page pp.pp.-
#Pages 6
Date of Issue