Efficient probabilistic image retrieval based on a mixed feature model

Authors:
Yi Zhuang;Zhiang Wu
Affiliations:
College of Computer and Information Engineering, Zhejiang Gongshang University, P.R. China;Jiangsu Provincial Key Laboratory of E-Business, Nanjing University of Finance and Economics, P.R. China
Venue:
APWeb'12 Proceedings of the 14th Asia-Pacific international conference on Web Technologies and Applications
Year:
2012

Citing 9
Cited 0

Photobook: content-based manipulation of image databases

International Journal of Computer Vision
Searching in high-dimensional spaces: Index structures for improving the performance of multimedia databases

ACM Computing Surveys (CSUR)
R-trees: a dynamic index structure for spatial searching

SIGMOD '84 Proceedings of the 1984 ACM SIGMOD international conference on Management of data
Query by Image and Video Content: The QBIC System

Computer
A Quantitative Analysis and Performance Study for Similarity-Search Methods in High-Dimensional Spaces

VLDB '98 Proceedings of the 24rd International Conference on Very Large Data Bases
The X-tree: An Index Structure for High-Dimensional Data

VLDB '96 Proceedings of the 22th International Conference on Very Large Data Bases
Independent Quantization: An Index Compression Technique for High-Dimensional Data Spaces

ICDE '00 Proceedings of the 16th International Conference on Data Engineering
iDistance: An adaptive B+-tree based indexing method for nearest neighbor search

ACM Transactions on Database Systems (TODS)
Toward Efficient Multifeature Query Processing

IEEE Transactions on Knowledge and Data Engineering

Quantified Score

Hi-index	0.00

Visualization

Abstract

In content-based image retrieval (CBIR) methods, color, texture, and shape are extracted from an image as low-level visual features for searching. However, these visual features could not express image sentiment concepts required in applications like garment search. To address this issue, we propose an efficient image retrieval method based on combined visual and conceptual feature spaces. A unified similarity distance between two images is obtained by linearly concatenating similarity measures. To further improve the retrieval efficiency, we propose an efficient probabilistic indexing scheme called Mixed-Feature-based Probabilistic Tree(MFP-Tree) to facilitate the retrieval over a large image repository. Different from conventional image retrieval and indexing methods, which only adopt visual similarity as a query metric, our proposed retrieval algorithm allows users to choose among the above three kinds of features as query elements. Moreover, a probabilistic model is introduced to refine the retrieval result with confidence guarantee. Comprehensive experiments have testified the effectiveness and efficiency of our proposed retrieval and indexing method.