Advertisement

Multimedia Tools and Applications

, Volume 78, Issue 3, pp 2877–2895 | Cite as

Effective product quantization-based indexing for nearest neighbor search

  • Chih-Yi ChiuEmail author
  • Jih-Sheng Chiu
  • Sarawut Markchit
  • Sheng-Hao Chou
Article
  • 174 Downloads

Abstract

Product quantization is a widely used lossy compression technique that can generate high quantization levels by a compact codebook set. It has been conducted in cluster-based index structures, termed as product quantization-based indexing. In this paper, we propose a novel product quantization-based indexing method for approximate nearest neighbor search. Inspired by the study for learning to rank, a ranking scheme is presented to learn the weighting relation between query-dependent features. The clusters in an index table are ranked by the relevance scores derived from the weighted features with respect to the query. We then present an approximate nearest neighbor search algorithm integrating the proposed ranking scheme with the product quantization-based index structure. Experimental results on the billion-level datasets demonstrate the effectiveness and superiority of the proposed method compared with several state-of-the-art methods.

Keywords

Approximate nearest neighbor search Product quantization Learning to rank Deep neural networks 

References

  1. 1.
    Babenko A, Lempitsky V (2014) Additive quantization for extreme vector compression. Proc IEEE Int Conf Comput Vis Pattern Recognit:931–938Google Scholar
  2. 2.
    Babenko A, Lempitsky V (2015) The inverted multi-index. IEEE Trans Pattern Anal Mach Intell 37(6):1247–1260CrossRefGoogle Scholar
  3. 3.
    Babenko A, Lempitsky V (2016) Efficient indexing of billion-scale datasets of deep descriptors. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2055–2063, Las Vegas, USA, Jun. 26-Jul. 1, 2016Google Scholar
  4. 4.
    Calonder M, Lepetit V, Strecha C, Fua P (2010) BRIEF: binary robust independent elementary features. Proc Eur conf Comput Vis:778–792Google Scholar
  5. 5.
    Chen Y, Guan T, Wang C (2010) Approximate nearest neighbor search by residual vector quantization. Sensors 10:11259–11273CrossRefGoogle Scholar
  6. 6.
    Chiu CY, Liou YC, Prayoonwong A (2016) Approximate asymmetric search for binary embedding codes. ACM Trans Multimed Comput Commun Appl 13(1):1–25CrossRefGoogle Scholar
  7. 7.
    Dai Q, Li J, Wang J, Jiang YG (2016) Binary optimized hashing. In Proceedings of ACM International Conference on Multimedia (SIGMM), pp. 1247–1256, Amsterdam, Netherlands, Oct. 15–19, 2016Google Scholar
  8. 8.
    Dong W, Charikar M, Li K (2008) Asymmetric distance estimation with sketches for similarity search in high-dimensional spaces. In Proceedings of ACM International Conference on Information Retrieval (SIGIR), pp. 128–130, 2008Google Scholar
  9. 9.
    Ge T, He K, Ke Q, Sun J (2013) Optimized product quantization for approximate nearest neighbor search. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Portland, USA, Jun. 23–28, 2013Google Scholar
  10. 10.
    Ge T, He K, Sun J (2014) graph cuts for supervised binary coding. Proc Eur conf Comput Vis 7:250–264Google Scholar
  11. 11.
    Gong Y, Lazebnik S (2011) iterative quantization: a procrustean approach to learning binary codes. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), 2011Google Scholar
  12. 12.
    Gordo A, Perronnin F, Gong Y, Lazebnik S (2014) Asymmetric distances for binary embeddings. IEEE Trans Pattern Anal Mach Intell 36(1):33–47CrossRefGoogle Scholar
  13. 13.
    He K, Wen F, Sun J. (2013) K-means hashing: an affinity-preserving quantization method for learning binary compact codes. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Portland, USA, Jun. 23–28, 2013Google Scholar
  14. 14.
    Indyk P, Motwani R (1998) Approximate nearest neighbors: towards removing the curse of dimensionality. In Proceedings of ACM Symposium on Theory of Computing (STOC), pp 604–613, 1998Google Scholar
  15. 15.
    Jégou H, Douze M, Schmid C (2011) Product quantization for nearest neighbor search. IEEE Trans Pattern Anal Mach Intell 33(1):2481–2488CrossRefGoogle Scholar
  16. 16.
    Jégou H, Tavenard R, Douze M, Amsaleg L (2011) Searching in one billion vectors: re-rank with source coding. In Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 861–864, Prague, Czech Republic, May 22–27, 2011Google Scholar
  17. 17.
    Kalantidis Y, Avrithis Y (2014) Locally optimized product quantization for approximate nearest neighbor search. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Columbus, USA, Jun. 23–28, 2014Google Scholar
  18. 18.
    Li J, Lan X, Wang J, Yang M, Zheng N (2017) Fast additive quantization for vector compression in nearest neighbor search. Multimed Tools Appl 76(22):23273–23289CrossRefGoogle Scholar
  19. 19.
    Liong VE, Lu J, Wang G, Moulin P, Zhou J (2015) Deep hashing for compact binary codes learning. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2475–2483, Boston, USA, Jun. 7–12, 2015Google Scholar
  20. 20.
    Liu W, Wang J, Ji R, Jiang YG, Chang SF (2012) Supervised hashing with kernels. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2123–2130, Rhode Island, USA, Jun. 16–21, 2012Google Scholar
  21. 21.
    Lowe DG (2004) Distinctive image features from scale-invariant keypoints. Int J Comput Vis 60(2):91–110MathSciNetCrossRefGoogle Scholar
  22. 22.
    Martinez J, Clement J, Hoos HH, Little JJ (2016) Revisiting additive quantization. In Proceedings of European Conference on Computer Vision (ECCV), pp. 137–153, Amsterdam, The Netherlands, Oct. 8–16, 2016Google Scholar
  23. 23.
    Matsui Y, Yamasaki T, Aizawa K (2015) PQTable: fast exact asymmetric distance neighbor search for product quantization using hash tables. In Proceedings of IEEE international conference on computer vision (ICCV), pp. 1940–1948, Santiago, Chile, 2015Google Scholar
  24. 24.
    Muja M, Lowe DG (2014) Scalable nearest neighbor algorithms for dimensional data. IEEE Trans Pattern Anal Mach Intell 36(11):2227–2240CrossRefGoogle Scholar
  25. 25.
    Nistér D, Stewénius H (2006) Scalable recognition with a vocabulary tree. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR). New York, USA, 2006Google Scholar
  26. 26.
    Norouzi M, Fleet DJ (2011) Minimal loss hashing for compact binary codes. In Proceedings of International Conference on Machine Learning (ICML), Bellevue, USA, Jun. 28-Jul. 2, 2011Google Scholar
  27. 27.
    Norouzi M, Fleet DJ (2013) Cartesian k-means. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Portland, USA, Jun. 23–28, 2013Google Scholar
  28. 28.
    Paulevé L, Jégou H, Amsaleg L (2010) Locality sensitive hashing: a comparison of hash function types and querying mechanisms. Pattern Recogn Lett 31:1348–1358CrossRefGoogle Scholar
  29. 29.
    Silpa-Anan C, Hartley R (2008) Optimized KD-trees for fast image descriptor matching. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Anchorage, USA, 2008Google Scholar
  30. 30.
    Torralba A, Fergus R, Weiss Y (2008) small codes and large image databases for recognition. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), 2008Google Scholar
  31. 31.
    Wang J, Kumar S, Chang SF (2010) semi-supervised hashing for scalable image retrieval. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), 2010Google Scholar
  32. 32.
    Wang J, Shen HT, Song J, Ji J (2014) Hashing for similarity search: a survey. ArXiv:1408.2927Google Scholar
  33. 33.
    Wang J, Shen HT, Yan S, Yu N, Li S, Wang J (2014) optimized distances for binary code ranking. In Proceedings of ACM international conference on multimedia (ACMMM), pp. 517–526, 2014Google Scholar
  34. 34.
    Wei B, Guan T, Yu J (2014) Projected residual vector quantization for ANN search. IEEE Multimedia 21(3):41–51CrossRefGoogle Scholar
  35. 35.
    Weiss Y, Fergus R, Torralba A (2012) Multi-dimensional spectral hashing. Proc Eur Conf Comput Vis 5:340–353Google Scholar
  36. 36.
    Xia Y, He K, Wen F, Sun J (2013) Joint inverted indexing. In Proceedings of IEEE International Conference on Computer Vision (ICCV), pp. 3416–3423, Sydney, Australia, Dec. 1–8, 2013Google Scholar
  37. 37.
    Zhang T, Du C, Wang J (2014) composite quantization for approximate nearest neighbor search. In Proceedings of International conference on Machine Learning (ICML), Beijing, China, Jun. 21–26, 2014Google Scholar
  38. 38.
    Zhang T, Qi GJ, Tang J, Wang J (2015) Sparse composite quantization. In Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), pp. 4548–4556, Boston, USA, Jun. 7–12, 2015Google Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  • Chih-Yi Chiu
    • 1
    Email author
  • Jih-Sheng Chiu
    • 1
  • Sarawut Markchit
    • 1
  • Sheng-Hao Chou
    • 1
  1. 1.Department of Computer Science and Information EngineeringNational Chiayi UniversityChiayi CityTaiwan

Personalised recommendations