![](https://d3ilqtpdwi981i.cloudfront.net/Xg_NAX_X7G4NLLbpYBK6lXDllrM=/425x550/smart/https://bepress-attached-resources.s3.amazonaws.com/uploads/a1/0a/4d/a10a4de2-1f25-4ec0-af72-ac542475de95/thumbnail_BPFile%20object.jpg)
Unpublished Paper
Statistical Models for Automatic Video Annotation and Retrieval
(2004)
Abstract
We apply a continuous relevance model (CRM) to the problem of directly retrieving the visual content of videos using text queries. The model computes a joint probability model for image features and words using a training set of annotated images. The model may then be used to annotate unseen test images. The probabilistic annotations are used for retrieval using text queries. We also propose a modified model - the normalized CRM - which substantially improves performance on a subset of the TREC Video dataset.
Disciplines
Publication Date
2004
Comments
This is the pre-published version harvested from CIIR.
Citation Information
V. Lavrenko, S. L. Feng and R. Manmatha. "Statistical Models for Automatic Video Annotation and Retrieval" (2004) Available at: http://works.bepress.com/r_manmatha/26/