A Bayesian Hierarchical Model for Learning Natural Scene Categories

We propose a novel approach to learn and recognize natural scene categories. Unlike previous work [9, 17], it does not require experts to annotate the training set. We represent the image of a scene by a collection of local regions, denoted as codewords obtained by unsupervised learning. Each region is represented as part of a "theme". In previous work, such themes were learnt from hand-annotations of experts, while our method learns the theme distributions as well as the codewords distribution over the themes without supervision. We report satisfactory categorization performances on a large set of 13 categories of complex scenes.

• Type: Conference
• Year: 2005
• Where: CVPR
• Authors: Fei-Fei Li 0002, Pietro Perona, California Institute of Technology
• Keywords: Computer Vision, CVPR 2005, Natural Scene Categories, Satisfactory Categorization Performances, Theme Distributions, Unsupervised Learning