ABSTRACT
The explosive growth of multimedia data poses serious challenges to data storage, management and search. Efficient near-duplicate detection is one of the required technologies for various applications. In this paper, we introduce MyFinder, an image near-duplicate detection system for large image collections. MyFinder consists of three major components: 1) a local-feature-based image representation utilizing the proposed LDP (Local-Difference-Pattern) feature, 2) the Locality-Sensitive-Hashing (LSH) as the core indexing structure to assure the most frequent data access occurred in the main memory, and 3) multi-step verification for queries to best exclude false positives and to increase the precision.
- Y. Ke, R. Sukthankar, and L. Huston. Efficient Near-duplicate Detection and Sub-image Retrieval. In Proceedings of ACM International Conference on Multimedia, 2004, pages 869--876. Google ScholarDigital Library
- A. Gionis, P. Indyk, and R. Motwani. Similarity search in high dimensions via hashing. In Proceedings of International Conference on Very Large Databases, 1999. Google ScholarDigital Library
- CGFA - A Virtual Art Museum. http://cgfa.sunsite.dk/Google Scholar
- D. Lowe. Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision, 2004 Google ScholarDigital Library
- B. Thomee, M. J. Huiskes, E. M. Bakker, and Michael S. Lew. Large scale image copy detection evaluation. In Proceedings of ACM International Conference on Multimedia Information Retrieval, 2008, pages 59--66 Google ScholarDigital Library
Index Terms
- MyFinder: near-duplicate detection for large image collections
Recommendations
Review of medical image retrieval systems and future directions
CBMS '11: Proceedings of the 2011 24th International Symposium on Computer-Based Medical SystemsThis paper presents a review of online systems for content-based medical image retrieval (CBIR). The objective of this review is to evaluate the capabilities and gaps in these systems and to determine ways of improving relevance of multi-modal (text and ...
Content-Based 3D Object Retrieval
Methods for automatically extracting descriptors from 3D objects are key to searching and indexing techniques in their growing repositories. The authors present two recently proposed approaches and discuss methods for benchmarking the 3D retrieval ...
Comments