Beyond "project and sign" for cosine estimation with binary codes - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2014

Beyond "project and sign" for cosine estimation with binary codes

Raghavendran Balu
Teddy Furon
Hervé Jégou
  • Function : Author
  • PersonId : 833473

Abstract

Many nearest neighbor search algorithms rely on encoding real vectors into binary vectors. The most common strategy projects the vectors onto random directions and takes the sign to produce so-called sketches. This paper discusses the sub-optimality of this choice, and proposes a better encoding strategy based on the quantization and reconstruction points of view. Our second contribution is a novel asymmetric estimator for the cosine similarity. Similar to previous asymmetric schemes, the query is not quantized and the similarity is computed in the compressed domain. Both our contribution leads to improve the quality of nearest neighbor search with binary codes. Its efficiency compares favorably against a recent encoding technique.
Fichier principal
Vignette du fichier
icassp_beyond_project_sign.pdf (198.44 Ko) Télécharger le fichier
Vignette du fichier
thumb.png (16.58 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Format : Figure, Image
Loading...

Dates and versions

hal-00942075 , version 1 (04-02-2014)

Identifiers

  • HAL Id : hal-00942075 , version 1

Cite

Raghavendran Balu, Teddy Furon, Hervé Jégou. Beyond "project and sign" for cosine estimation with binary codes. ICASPP - International Conference on Acoustics, Speech, and Signal Processing, IEEE, May 2014, Florence, Italy. ⟨hal-00942075⟩
433 View
507 Download

Share

Gmail Facebook X LinkedIn More