K-SUBSPACES QUANTIZATION FOR APPROXIMATE NEAREST NEIGHBOR SEARCH
0 likes113 views
This document presents a novel vector quantization method for approximate nearest neighbor (ANN) search that enhances the speed and accuracy of data retrieval. The method utilizes iterative techniques to minimize quantization errors in affine subspaces, demonstrating superior performance compared to existing algorithms while maintaining similar computational costs. Future work includes extending the approach for non-exhaustive searches and testing on labeled datasets.
![CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249)
MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com
Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com
outperform the state-of-the-art-methods, with comparable computational cost
and additional storage. In this paper it is also shown that, dimension reduction
is an important source of quantization error, and by exploiting subspace
clustering techniques the quantization error can be reduced, leading to a
better quantization performance. So far we have focused mainly on exhaustive
search but an index-based non-exhaustive extension for the propose method
can be further investigated. Our approach can also be extended to labeled
datasets in order to test k-nearest neighbor classification performance. These
will be the topics of our future work.
REFERENCES
[1] P. Indyk and R. Motwani, “Approximate nearest neighbors: Towards
removing the curse of dimensionality,” in Proc. 30th Annu. ACM Symp. Theory
Comput., 1998, pp. 604–613.
[2] J. Wang, H. T. Shen, J. Song, and J. Ji, “Hashing for similarity search: A
survey,” arXiv preprint, 2014, p. 1408.2927.
[3] M. Datar, N. Immorlica, P. Indyk, and V. S. Mirrokni, “Localitysensitive
hashing scheme based on P-stable distributions,” in Proc. 20th Annu. Symp.
Comput. Geom., 2004, pp. 253–262.
[4] K. Terasawa and Y. Tanaka, “Spherical LSH for approximate nearest
neighbor search on unit hypersphere,” in Proc. 10th Int. Conf. Algorithms Data
Struct., 2007, pp. 27–38.](https://image.slidesharecdn.com/k-subspacesquantizationforapproximated-160722070942/85/K-SUBSPACES-QUANTIZATION-FOR-APPROXIMATE-NEAREST-NEIGHBOR-SEARCH-2-320.jpg)
![CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249)
MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com
Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com
[5] X. He, D. Cai, S. Yan, and H. Zhang, “Neighborhood preserving embedding,”
in Proc. 10th IEEE Int. Conf. Comput. Vis., 2005, pp. 1208–1213.
[6] H. Jegou, M. Douze, C. Schmid, and P. Perez, “Aggregating local descriptors
into a compact image representation,” in Proc. IEEE Conf. Comput. Vis. Pattern
Recog., 2010, pp. 3304–3311.
[7] J. Heo, Y. Lee, and J. He, “Spherical hashing,” in Proc. IEEE Conf. Comput.
Vis. Pattern Recog., 2012, pp. 2957–2964.
[8] A. Gordo, F. Perronnin, Y. Gong, and S. Lazebnik, “Asymmetric distances for
binary embeddings,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 36, no. 1, pp.
33–47, Jan. 2014.
[9] W. Dong, M. Charikar, and K. Li, “Asymmetric distance estimation with
sketches for similarity search in high-dimensional spaces,” in Proc. 31st Annu.
Int. ACM SIGIR Conf. Res. Develop. Inf. Retrieval, 2008, pp. 123–130.
[10] S. Lloyd, “Least squares quantization in PCM,” IEEE Trans. Inf. Theory, vol.
28, no. 2, pp. 129–137, Mar. 1982.](https://image.slidesharecdn.com/k-subspacesquantizationforapproximated-160722070942/85/K-SUBSPACES-QUANTIZATION-FOR-APPROXIMATE-NEAREST-NEIGHBOR-SEARCH-3-320.jpg)
K-SUBSPACES QUANTIZATION FOR APPROXIMATE NEAREST NEIGHBOR SEARCH
- 1. CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249) MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com K-SUBSPACES QUANTIZATION FOR APPROXIMATE NEAREST NEIGHBOR SEARCH. Abstract Approximate Nearest Neighbor (ANN) search has become a popular approach for performing fast and efficient retrieval on very large-scale datasets in recent years, as the size and dimension of data grow continuously. In this paper, we propose a novel vector quantization method for ANN search which enables faster and more accurate retrieval on publicly available datasets. We define vector quantization as a multiple affine subspace learning problem and explore the quantization centroids on multiple affine subspaces. We propose an iterative approach to minimize the quantization error in order to create a novel quantization scheme, which outperforms the state-of-the-art algorithms. The computational cost of our method is also comparable to that of the competing methods. CONCLUSION In this study a novel vector quantization algorithm is proposed for the approximate nearest neighbor search problem. The proposed method explores the quantization centers in affine subspaces through an iterative technique, which jointly attempts to minimize the quantization error of the training samples in the learnt subspaces, while minimizing the projection error of the samples to the corresponding subspaces. The proposed method has proven to
- 2. CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249) MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com outperform the state-of-the-art-methods, with comparable computational cost and additional storage. In this paper it is also shown that, dimension reduction is an important source of quantization error, and by exploiting subspace clustering techniques the quantization error can be reduced, leading to a better quantization performance. So far we have focused mainly on exhaustive search but an index-based non-exhaustive extension for the propose method can be further investigated. Our approach can also be extended to labeled datasets in order to test k-nearest neighbor classification performance. These will be the topics of our future work. REFERENCES [1] P. Indyk and R. Motwani, “Approximate nearest neighbors: Towards removing the curse of dimensionality,” in Proc. 30th Annu. ACM Symp. Theory Comput., 1998, pp. 604–613. [2] J. Wang, H. T. Shen, J. Song, and J. Ji, “Hashing for similarity search: A survey,” arXiv preprint, 2014, p. 1408.2927. [3] M. Datar, N. Immorlica, P. Indyk, and V. S. Mirrokni, “Localitysensitive hashing scheme based on P-stable distributions,” in Proc. 20th Annu. Symp. Comput. Geom., 2004, pp. 253–262. [4] K. Terasawa and Y. Tanaka, “Spherical LSH for approximate nearest neighbor search on unit hypersphere,” in Proc. 10th Int. Conf. Algorithms Data Struct., 2007, pp. 27–38.
- 3. CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249) MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com [5] X. He, D. Cai, S. Yan, and H. Zhang, “Neighborhood preserving embedding,” in Proc. 10th IEEE Int. Conf. Comput. Vis., 2005, pp. 1208–1213. [6] H. Jegou, M. Douze, C. Schmid, and P. Perez, “Aggregating local descriptors into a compact image representation,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog., 2010, pp. 3304–3311. [7] J. Heo, Y. Lee, and J. He, “Spherical hashing,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog., 2012, pp. 2957–2964. [8] A. Gordo, F. Perronnin, Y. Gong, and S. Lazebnik, “Asymmetric distances for binary embeddings,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 36, no. 1, pp. 33–47, Jan. 2014. [9] W. Dong, M. Charikar, and K. Li, “Asymmetric distance estimation with sketches for similarity search in high-dimensional spaces,” in Proc. 31st Annu. Int. ACM SIGIR Conf. Res. Develop. Inf. Retrieval, 2008, pp. 123–130. [10] S. Lloyd, “Least squares quantization in PCM,” IEEE Trans. Inf. Theory, vol. 28, no. 2, pp. 129–137, Mar. 1982.