Scene Parsing Using Region-Based Generative Models

  • Authors:
  • M. R. Boutell;J. Luo;C. M. Brown

  • Affiliations:
  • Dept. of Comput. Sci. & Software Eng., Rose-Hulman Inst. of Technol., Terre Haute, IN;-;-

  • Venue:
  • IEEE Transactions on Multimedia
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Semantic scene classification is a challenging problem in computer vision. In contrast to the common approach of using low-level features computed from the whole scene, we propose "scene parsing" utilizing semantic object detectors (e.g., sky, foliage, and pavement) and region-based scene-configuration models. Because semantic detectors are faulty in practice, it is critical to develop a region-based generative model of outdoor scenes based on characteristic objects in the scene and spatial relationships between them. Since a fully connected scene configuration model is intractable, we chose to model pairwise relationships between regions and estimate scene probabilities using loopy belief propagation on a factor graph. We demonstrate the promise of this approach on a set of over 2000 outdoor photographs, comparing it with existing discriminative approaches and those using low-level features