Semantic image retrieval based on probabilistic latent semantic analysis
Document Type
Conference Proceeding
Publication Date
12-1-2006
Abstract
Content-based image retrieval (CBIR) systems combine computer vision techniques and learning methodologies to find images in the database similar to the query images. Relevance feedback methods are introduced to the CBIR area as a tool to help the user to guide the retrieval system during the search process. Search history of the retrieval system, which is the accumulated feedbacks from past retrievals, has been recently used as a prior knowledge to improve the image retrieval performance. In this paper, we introduce an image retrieval model based on probabilistic latent semantic analysis (PLSA) that utilizes the system's search history to find hidden image semantics of the database. Image features are integrated to the model as well. The model is capable of detecting images and image features that efficiently represent semantic classes in the database. We demonstrate the effectiveness of our approach by comparing to previous work in this area.
Publication Source (Journal or Book title)
Proceedings of the 14th Annual ACM International Conference on Multimedia, MM 2006
First Page
703
Last Page
706
Recommended Citation
Shah-Hosseini, A., & Knapp, G. (2006). Semantic image retrieval based on probabilistic latent semantic analysis. Proceedings of the 14th Annual ACM International Conference on Multimedia, MM 2006, 703-706. https://doi.org/10.1145/1180639.1180788