Parameter Estimation and Energy Minimization for Region-based Semantic Segmentation

Abstract : We consider the problem of parameter estimation and energy minimization for a region-based semantic segmentation model. The model divides the pixels of an image into non-overlapping connected regions, each of which is assigned a label indicating its semantic class. In the context of energy minimization, the main problem we face is the large number of putative pixel-to-region assignments. We address this problem by designing an accurate linear programming based approach for selecting the best set of regions from a large dictionary. The dictionary is constructed by merging and intersecting segments obtained from multiple bottom-up over-segmentations. The linear program is solved efficiently using dual decomposition. In the context of parameter estimation, the main problem we face is the lack of fully supervised data. We address this issue by developing a principled framework for parameter estimation using diverse data. More precisely, we propose a latent structural support vector machine formulation, where the latent variables model any missing information in the human annotation. Of particular interest to us are three types of annotations: (i) images segmented using generic foreground or background classes; (ii) images with bounding boxes specified for objects; and (iii) images labeled to indicate the presence of a class. Using large, publicly available datasets we show that our methods are able to significantly improve the accuracy of the region-based model.
Complete list of metadatas

Cited literature [44 references]  Display  Hide  Download
Contributor : Puneet Kumar Dokania <>
Submitted on : Wednesday, September 4, 2013 - 11:28:13 AM
Last modification on : Thursday, February 7, 2019 - 2:00:01 PM
Long-term archiving on : Thursday, December 5, 2013 - 4:16:42 AM


Files produced by the author(s)


  • HAL Id : hal-00857918, version 1



M. Pawan Kumar, Haithem Turki, Dan Preston, Daphne Koller. Parameter Estimation and Energy Minimization for Region-based Semantic Segmentation. [Technical Report] 2013. ⟨hal-00857918⟩



Record views


Files downloads