Abstract
Retrieval techniques based on pure similarity metrics are often suffered from the scales of image features. An alternative approach is to learn a mapping based on queries and relevance feedback by supervised learning. However, the learning is plagued by the insufficiency of labeled training images. Different from most current research in image retrieval, this paper investigates the possibility of taking advantage of unlabeled images in the given image database to make a hybrid statistical learning feasible. Assuming a generative model of the database, the proposed approach casts image retrieval as a transductive learning problem in a probabilistic framework. Our experiments show that the proposed approach has a satisfactory performance in image retrieval applications.
Original language | English (US) |
---|---|
Title of host publication | Proceedings - 15th International Conference on Pattern Recognition, ICPR 2000 - Volume 1 |
Subtitle of host publication | Computer Vision and Image Analysis |
Publisher | Institute of Electrical and Electronics Engineers Inc. |
Pages | 21-24 |
Number of pages | 4 |
ISBN (Electronic) | 0769507506 |
DOIs | |
State | Published - 2000 |
Event | 15th International Conference on Pattern Recognition, ICPR 2000 - Barcelona, Spain Duration: Sep 3 2000 → Sep 7 2000 |
Publication series
Name | Proceedings - International Conference on Pattern Recognition |
---|---|
Volume | 1 |
ISSN (Print) | 1051-4651 |
Other
Other | 15th International Conference on Pattern Recognition, ICPR 2000 |
---|---|
Country/Territory | Spain |
City | Barcelona |
Period | 9/3/00 → 9/7/00 |
Funding
This work was supported in part by National Science Foundation Grants CDA-96-24396, IRI-96-34618 and EIA-99-75019.
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition