Learning specific-class segmentation from diverse data
Manish Kumar, Haithem Turki, Dan Preston et al. · International Conference on Computer Vision · 2011
We consider the task of learning the parameters of a segmentation model that assigns a specific semantic class to each pixel of a given image. The main problem we face is the lack of fully supervised data. We address this issue by developing a principled framework for learning the parameters of a specific-class segmentation model 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 approach is able to exploit the information present in different annotations to improve the accuracy of a state-of-the art region-based model.
Cited by 77 works. doi:10.1109/iccv.2011.6126446
Authors
References
- ImageNet: A large-scale hierarchical image database
- The Pascal Visual Object Classes (VOC) Challenge
- A discriminatively trained, multiscale, deformable part model
- TextonBoost: Joint Appearance, Shape and Context Modeling for Multi-class Object Recognition and Segmentation
- Multiscale conditional random fields for image labeling
- Pegasos
- Decomposing a scene into geometric and semantically consistent regions
- ClassCut for Unsupervised Class Segmentation
- Pegasos: primal estimated sub-gradient solver for SVM
- LOCUS: learning object classes with unsupervised segmentation