Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
A Compositional Exemplar-Based Model for Hair Segmentation
 

Summary: A Compositional Exemplar-Based Model
for Hair Segmentation
Nan Wang1
, Haizhou Ai1
, and Shihong Lao2
1
Computer Science & Technology Department, Tsinghua University, Beijing, China
ahz@mail.tsinghua.edu.cn
2
Core Technology Center, Omron Corporation, Kyoto, Japan
lao@ari.ncl.omron.co.jp
Abstract. Hair is a very important part of human appearance. Robust
and accurate hair segmentation is difficult because of challenging vari-
ation of hair color and shape. In this paper, we propose a novel Com-
positional Exemplar-based Model (CEM) for hair style segmentation.
CEM generates an adaptive hair style (a probabilistic mask) for the in-
put image automatically in the manner of Divide-and-Conquer, which
can be divided into decomposition stage and composition stage natu-
rally. For the decomposition stage, we learn a strong ranker based on
a group of weak similarity functions emphasizing the Semantic Layout

  

Source: Ai, Haizhou - Department of Computer Science and Technology, Tsinghua University

 

Collections: Computer Technologies and Information Sciences