1 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
2 TEXMEX - Multimedia content-based indexing
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
Abstract : It is well known that high-dimensional nearest-neighbor retrieval is very expensive. Many signal processing methods suffer from this computing cost. Dramatic performance gains can be obtained by using approximate search, such as the popular Locality-Sensitive Hashing. This paper improves LSH by performing an on-line selection of the most appropriate hash functions from a pool of functions. An additional improvement originates from the use of $E_8$ lattices for geometric hashing instead of one-dimensional random projections. A performance study based on state-of-the-art high-dimensional descriptors computed on real images shows that our improvements to LSH greatly reduce the search complexity for a given level of accuracy.
Document type :
Conference papers

https://hal.inria.fr/inria-00318614
Contributor : Hervé Jégou <>
Submitted on : Tuesday, March 15, 2011 - 2:51:35 PM
Last modification on : Monday, December 17, 2018 - 11:22:02 AM
Document(s) archivé(s) le : Thursday, November 8, 2012 - 11:45:18 AM

### Files

qalsh.pdf
Files produced by the author(s)

### Citation

Hervé Jégou, Laurent Amsaleg, Cordelia Schmid, Patrick Gros. Query-Adaptative Locality Sensitive Hashing. ICASSP 2008 - IEEE International Conference on Acoustics, Speech, and Signal Processing, Mar 2008, Las Vegas, United States. pp.825-828, ⟨10.1109/ICASSP.2008.4517737⟩. ⟨inria-00318614⟩

Record views