Modeling annotated data
- 28 July 2003
- conference paper
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 127-134
- https://doi.org/10.1145/860435.860460
Abstract
We consider the problem of modeling annotated data---data with multiple types where the instance of one type (such as a caption) serves as a description of the other type (such as an image). We describe three hierarchical probabilistic mixture models which aim to describe such data, culminating in correspondence latent Dirichlet allocation, a latent variable model that is effective at modeling the joint distribution of both types and the conditional distribution of the annotation given the primary type. We conduct experiments on the Corel database of images and captions, assessing performance in terms of held-out likelihood, automatic annotation, and text-based image retrieval.Keywords
This publication has 8 references indexed in Scilit:
- Automatic image annotation and retrieval using cross-media relevance modelsPublished by Association for Computing Machinery (ACM) ,2003
- A model of multimedia information retrievalJournal of the ACM, 2001
- A probabilistic framework for semantic video indexing, filtering, and retrievalIEEE Transactions on Multimedia, 2001
- Normalized cuts and image segmentationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2000
- Image Information Retrieval: An Overview of Current ResearchInforming Science: The International Journal of an Emerging Transdiscipline, 2000
- An Introduction to Variational Methods for Graphical ModelsMachine Learning, 1999
- A language modeling approach to information retrievalPublished by Association for Computing Machinery (ACM) ,1998
- Parametric Empirical Bayes Inference: Theory and ApplicationsJournal of the American Statistical Association, 1983