Conditional random fields for image region labeling with global observation

Zhe Lin, Wen Chan, Kai He, Xiangdong Zhou, Mei Wang

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

1 Citation (Scopus)


Region (pixel) labeling has attracted increasing attentions from both research and industry communities. In this paper, we present a new approach based on Conditional Random Fields (CRF) to assign the semantic labels to the corresponding regions of images. Different from previous work, our model incorporates the global observation into the region labeling framework with the harness of spatial context modeling of CRF model. The experimental results with two commonly used datasets demonstrate that our method achieves significant improvement on region labeling tasks compared with the strong baselines.

Original languageEnglish
Title of host publicationAdvances in Multimedia Information Processing, PCM 2013 - 14th Pacific-Rim Conference on Multimedia, Proceedings
PublisherSpringer Verlag
Number of pages12
ISBN (Print)9783319037301
Publication statusPublished - 2013
Externally publishedYes
Event14th Pacific-Rim Conference on Multimedia, PCM 2013 - Nanjing, China
Duration: 13 Dec 201316 Dec 2013

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8294 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference14th Pacific-Rim Conference on Multimedia, PCM 2013


  • Conditional random fields
  • Global observation
  • Image annotation
  • Image region labeling


Dive into the research topics of 'Conditional random fields for image region labeling with global observation'. Together they form a unique fingerprint.

Cite this