Learning Dictionary of Discriminative Part Detectors for Image Categorization and Cosegmentation

This paper proposes a novel approach to learning mid-level image models for image categorization and cosegmentation. We represent each image class by a dictionary of part detectors that best discriminate that class from the background. We learn category-specific part detectors in a weakly supervised setting in which the training images are only annotated with category labels without part/object location information. We use a latent SVM model regularized using the ℓ2,1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\ell _{2,1}$$\end{document} group sparsity norm to learn the part detectors. Starting from a large set of initial parts, the group sparsity regularizer forces the model to jointly select and optimize a set of discriminative part detectors in a max-margin framework. We propose a stochastic version of a proximal algorithm to solve the corresponding optimization problem. We apply the learned part detectors to image classification and cosegmentation, and present extensive comparative experiments with standard benchmarks.