Skip to main content
SpringerLink
Log in

Dynamic behavior of balanced NV-trees

  • Regular Paper
  • Published:
Multimedia Systems Aims and scope Submit manuscript

Abstract

In recent years, some approximate high-dimensional indexing techniques have shown promising results by trading off quality guarantees for improved query performance. While the query performance and quality of these methods has been well studied, however, the performance of index maintenance has not yet been reported in any detail. Here, we focus on the dynamic behavior of the balanced NV-tree, which is a disk-based approximate index for very large collections. We report on an initial study of the effects of several implementation choices for the balanced NV-tree, and show that with appropriate implementation, significant performance improvements are possible. Overall, the proposed techniques not only reduce maintenance cost, but can also improve search performance significantly with minimal loss of search quality.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16

Similar content being viewed by others

Notes

  1. Since the split partition contained at least p descriptors, as many as p random disk reads may be required to find the descriptors. Unless the collection is very large, p random reads cost far more than a sequential scan.

  2. Note that the lack of quality in single index searches is due to the index configuration used, which is a balanced NV-tree with leaves of 32 pages; this configuration was chosen since it generates small indices very quickly. In [8], it was shown that with smaller leaves, better line selections and other configurations, the unbalanced NV-tree is quite effective for single index searches.

References

  1. Amsaleg, L., Gros, P.: Content-based retrieval using local descriptors: problems and issues from a database perspective. Pattern Anal. Appl. 4(2/3), 108–124 (2001)

    MATH  MathSciNet  Google Scholar 

  2. Bainbridge, D., Browne, P., Cairns, P., Rüger, S., Li-Qun, X.: Managing the growth of multimedia digital content. ERCIM News (62), 16–17 (2005)

  3. Casey, M., Slaney, M.: Song intersection by approximate nearest neighbor search. In: Proceedings of International Conference on Music Information Retrieval (ISMIR), Victoria, BC, Canada (2006)

  4. Chierichetti, F., Panconesi, A., Raghavan, P., Sozio, M., Tiberi, A., Upfal, E.: Finding near neighbors through cluster pruning. In: Proceedings of the ACM Symposium on Principles of Database Systems (PODS), Beijing, China (2007)

  5. Datar, M., Indyk, P., Immorlica, N., Mirrokni, V.S.: Locality-Sensitive Hashing using Stable Distributions. MIT Press, Cambridge (2006)

    Google Scholar 

  6. Fagin, R., Kumar, R., Sivakumar, D.: Efficient similarity search and classification via rank aggregation. In: Proceedings of the ACM SIGMOD International Conference on Management of Data. San Diego, CA, USA (2003)

  7. Lejsek, H., Ásmundsson, F.H., Jónsson, B.Þ., Amsaleg, L.: Scalability of local image descriptors: a comparative study. In: Proceedings of the ACM International Conference on Multimedia. Santa Barbara, CA, USA (2006)

  8. Lejsek, H., Ásmundsson, F.H., Jónsson, B.Þ., Amsaleg, L.: NV-tree: an efficient disk-based index for approximate search in very large high-dimensional collections. IEEE Trans. Pattern Anal. Mach. Intell. 31(5), 869–883 (2009)

    Article  Google Scholar 

  9. Li, C., Chang, E., Garcia-Molina, H., Wiederhold, G.: Clustering for approximate similarity search in high-dimensional spaces. IEEE Trans. Knowl. Data Eng. 14(4), 792–808 (2002)

    Article  Google Scholar 

  10. Liu, T., Moore, A., Gray, A., Yang, K.: An investigation of practical approximate nearest neighbor algorithms. In: Proceedings of Neural Information Processing Systems (NIPS). Vancouver, BC, Canada (2004)

  11. Lowe, D.G.: Distinctive image features from scale-invariant keypoints. Int. J. Comput. Vis. 60(2), (2004)

  12. Ólafsson, A.: Dynamic behavior of balanced NV-trees. Master’s Thesis, School of Computer Science, Reykjavík University (2008)

  13. Ólafsson, A., Jónsson, B.Þ., Amsaleg, L.: Dynamic behavior of balanced NV-trees. In: Proceedings of Content-Based Multimedia Indexing. London, England (2008)

  14. Shaft, U., Ramakrishnan, R.: Theory of nearest neighbors indexability. ACM Trans. Database Syst. 31(3), 814–838 (2006)

    Article  Google Scholar 

  15. Zhang, L., Tian, Q.: Multimedia content analysis: model-based approaches vs. data-driven approaches. In: Proceedings of the ACM International Conference on Multimedia. Beijing, China (2009)

Download references

Author information

Authors and Affiliations

  1. School of Computer Science, Reykjavík University, Reykjavík, Iceland

    Arnar Ólafsson & Björn Þór Jónsson

  2. IRISA–CNRS, Rennes, France

    Laurent Amsaleg

  3. Videntifier Technologies ehf., Reykjavík, Iceland

    Herwig Lejsek

Authors
  1. Arnar Ólafsson
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Björn Þór Jónsson
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Laurent Amsaleg
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Herwig Lejsek
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Laurent Amsaleg.

Additional information

Communicated by Balakrishnan Prabhakaran.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Ólafsson, A., Þór Jónsson, B., Amsaleg, L. et al. Dynamic behavior of balanced NV-trees. Multimedia Systems 17, 83–100 (2011). https://doi.org/10.1007/s00530-010-0199-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00530-010-0199-4

Keywords

Search

Navigation