A Statistical Model for General Contextual Object Recognition

A Statistical Model for General Contextual Object Recognition (2004)

Cached

Download Links

Other Repositories/Bibliography

DBLP

by Peter Carbonetto , Nando de Freitas , Kobus Barnard

Venue:	IN ECCV
Citations:	131 - 7 self

BibTeX

@INPROCEEDINGS{Carbonetto04astatistical,
    author = {Peter Carbonetto and Nando de Freitas and Kobus Barnard},
    title = {A Statistical Model for General Contextual Object Recognition},
    booktitle = {IN ECCV},
    year = {2004},
    pages = {350--362},
    publisher = {}
}

OpenURL

Abstract

We consider object recognition as the process of attaching meaningful labels to specific regions of an image, and propose a model that learns spatial relationships between objects. Given a set of images and their associated text (e.g. keywords, captions, descriptions), the objective is to segment an image, in either a crude or sophisticated fashion, then to find the proper associations between words and regions. Previous models are limited by the scope of the representation. In particular, they fail to exploit spatial context in the images and words. We develop a more expressive model that takes this into account. We formulate a spatially consistent probabilistic mapping between continuous image feature vectors and the supplied word tokens. By learning both word-to-region associations and object relations, the proposed model augments scene segmentations due to smoothing implicit in spatial consistency. Context introduces cycles to the undirected graph, so we cannot rely on a straightforward implementation of the EM algorithm for estimating the model parameters and densities of the unknown alignment variables. Instead, we develop an approximate EM algorithm that uses loopy belief propagation in the inference step and iterative scaling on the pseudo-likelihood approximation in the parameter update step. The experiments indicate that our approximate inference and learning algorithm converges to good local solutions. Experiments on a diverse array of images show that spatial context considerably improves the accuracy of object recognition. Most

Keyphrases

general contextual object recognition statistical model spatial context object recognition unknown alignment variable parameter update step model augments scene segmentation straightforward implementation proper association loopy belief propagation inference step continuous image feature vector supplied word token undirected graph consistent probabilistic mapping word-to-region association algorithm converges sophisticated fashion model parameter previous model specific region diverse array expressive model meaningful label em algorithm approximate inference pseudo-likelihood approximation iterative scaling spatial consistency approximate em algorithm object relation good local solution spatial relationship