Skip to main content
placeholder image

Efficient learning to label images

Conference Paper


Abstract


  • Conditional random field methods (CRFs) have gained popularity for image labeling tasks in recent years. In this paper, we describe an alternative discriminative approach, by extending the large margin principle to incorporate spatial correlations among neighboring pixels. In particular, by explicitly enforcing the sub modular condition, graph-cuts is conveniently integrated as the inference engine to attain the optimal label assignment efficiently. Our approach allows learning a model with thousands of parameters, and is shown to be capable of readily incorporating higher-order scene context. Empirical studies on a variety of image datasets suggest that our approach performs competitively compared to the state-of-the-art scene labeling methods.

Authors


  •   Jia, Ke (external author)
  •   Cheng, Li (external author)
  •   Liu, Nianjun (external author)
  •   Wang, Lei

Publication Date


  • 2010

Citation


  • Jia, K., Cheng, L., Liu, N. & Wang, L. (2010). Efficient learning to label images. Proceedings of Twentieth International Conference on Pattern Recognition (ICIP) (pp. 942-945). Australia: IEEE.

Scopus Eid


  • 2-s2.0-78149491985

Ro Metadata Url


  • http://ro.uow.edu.au/eispapers/532

Start Page


  • 942

End Page


  • 945

Abstract


  • Conditional random field methods (CRFs) have gained popularity for image labeling tasks in recent years. In this paper, we describe an alternative discriminative approach, by extending the large margin principle to incorporate spatial correlations among neighboring pixels. In particular, by explicitly enforcing the sub modular condition, graph-cuts is conveniently integrated as the inference engine to attain the optimal label assignment efficiently. Our approach allows learning a model with thousands of parameters, and is shown to be capable of readily incorporating higher-order scene context. Empirical studies on a variety of image datasets suggest that our approach performs competitively compared to the state-of-the-art scene labeling methods.

Authors


  •   Jia, Ke (external author)
  •   Cheng, Li (external author)
  •   Liu, Nianjun (external author)
  •   Wang, Lei

Publication Date


  • 2010

Citation


  • Jia, K., Cheng, L., Liu, N. & Wang, L. (2010). Efficient learning to label images. Proceedings of Twentieth International Conference on Pattern Recognition (ICIP) (pp. 942-945). Australia: IEEE.

Scopus Eid


  • 2-s2.0-78149491985

Ro Metadata Url


  • http://ro.uow.edu.au/eispapers/532

Start Page


  • 942

End Page


  • 945