Local Linear Spectral Hashing
Hashing for large scale image retrieval has become more and more popular because of its improvement in computational speed and storage reduction. Spectral Hashing (SH) is a very efficient unsupervised hashing method through mapping similar images to similar binary codes. However, it doesn’t take the non-neighbor points into consideration, and its assumption of uniform data distribution is usually not true. In this paper, we propose a \(local\ linear\ spectral\ hashing\) framework that minimizes the average Hamming distance with a new local neighbor matrix, which can guarantee the mapping not only from neighbor images to neighbor codes, but also from non-neighbor images to non-neighbor codes. Based on the framework, we utilize three linear methods to handle the proposed problem, including orthogonal hashing, non-orthogonal hashing, and sequential hashing. The experiments on two huge datasets demonstrate the efficiency and accuracy of our methods.
KeywordsImage Retrieval Hamming Distance Spectral Hashing Eigenvalue Decomposition
Unable to display preview. Download preview PDF.
- 1.Silpa-Anan, C., Hartley, R.: Optimised kd-trees for fast image descriptor matching. In: IEEE Conference on Computer Vision and Pattern Recognition, pp. 1–8 (2008)Google Scholar
- 3.Gionis, A., Indyk, P., Motwani, R., et al.: Similarity search in high dimensions via hashing. In: Proceedings of the International Conference on Very Large Data Bases, pp. 518–529 (1999)Google Scholar
- 4.Datar, M., Immorlica, N., Indyk, P., Mirrokni, V.S.: Locality-sensitive hashing scheme based on p-stable distributions. In: Proceedings of the Twentieth Annual Symposium on Computational Geometry, pp. 253–262. ACM (2004)Google Scholar
- 5.Kulis, B., Grauman, K.: Kernelized locality-sensitive hashing for scalable image search. In: IEEE 12th International Conference on Computer Vision, pp. 2130–2137 (2009)Google Scholar
- 6.Weiss, Y., Torralba, A., Fergus, R.: Spectral hashing. In: Advances in Neural Information Processing Systems, pp. 1753–1760 (2008)Google Scholar
- 8.Salakhutdinov, R., Hinton, G.: Learning a nonlinear embedding by preserving class neighbourhood structure. In: AI and Statistics, vol. 3, p. 5 (2007)Google Scholar