Joint K-Means Quantization for Approximate Nearest Neighbor Search

Ezgi Can Ozan; Serkan Kiranyaz; Moncef Gabbouj

doi:10.1109/ICPR.2016.7900200

Joint K-Means Quantization for Approximate Nearest Neighbor Search

Ezgi Can Ozan, Serkan Kiranyaz, Moncef Gabbouj

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

1 Citation (Scopus)

Abstract

Recently, Approximate Nearest Neighbor (ANN) Search has become a very popular approach for similarity search on large-scale datasets. In this paper, we propose a novel vector quantization method for ANN, which introduces a joint multi-layer K-Means clustering solution for determination of the codebooks. The performance of the proposed method is improved further by a joint encoding scheme. Experimental results verify the success of the proposed algorithm as it outperforms the state-of-the-art methods.

Original language	English
Title of host publication	23rd International Conference on Pattern Recognition (ICPR 2016)
Publisher	IEEE
ISBN (Electronic)	978-1-5090-4847-2
DOIs	https://doi.org/10.1109/ICPR.2016.7900200
Publication status	Published - 2017
Publication type	A4 Article in conference proceedings
Event	International Conference on Pattern Recognition - Duration: 1 Jan 1900 → …

Conference

Conference	International Conference on Pattern Recognition
Period	1/01/00 → …

Publication forum classification

Publication forum level 1

Access to Document

10.1109/ICPR.2016.7900200

Cite this

@inproceedings{47e5820fe98d499a83d81ea2c98fdc9f,

title = "Joint K-Means Quantization for Approximate Nearest Neighbor Search",

abstract = "Recently, Approximate Nearest Neighbor (ANN) Search has become a very popular approach for similarity search on large-scale datasets. In this paper, we propose a novel vector quantization method for ANN, which introduces a joint multi-layer K-Means clustering solution for determination of the codebooks. The performance of the proposed method is improved further by a joint encoding scheme. Experimental results verify the success of the proposed algorithm as it outperforms the state-of-the-art methods.",

author = "Ozan, {Ezgi Can} and Serkan Kiranyaz and Moncef Gabbouj",

note = "JUFOID=58099 EXT={"}Kiranyaz, Serkan{"}; International Conference on Pattern Recognition ; Conference date: 01-01-1900",

year = "2017",

doi = "10.1109/ICPR.2016.7900200",

language = "English",

booktitle = "23rd International Conference on Pattern Recognition (ICPR 2016)",

publisher = "IEEE",

}

TY - GEN

T1 - Joint K-Means Quantization for Approximate Nearest Neighbor Search

AU - Ozan, Ezgi Can

AU - Kiranyaz, Serkan

AU - Gabbouj, Moncef

N1 - JUFOID=58099 EXT="Kiranyaz, Serkan"

PY - 2017

Y1 - 2017

N2 - Recently, Approximate Nearest Neighbor (ANN) Search has become a very popular approach for similarity search on large-scale datasets. In this paper, we propose a novel vector quantization method for ANN, which introduces a joint multi-layer K-Means clustering solution for determination of the codebooks. The performance of the proposed method is improved further by a joint encoding scheme. Experimental results verify the success of the proposed algorithm as it outperforms the state-of-the-art methods.

AB - Recently, Approximate Nearest Neighbor (ANN) Search has become a very popular approach for similarity search on large-scale datasets. In this paper, we propose a novel vector quantization method for ANN, which introduces a joint multi-layer K-Means clustering solution for determination of the codebooks. The performance of the proposed method is improved further by a joint encoding scheme. Experimental results verify the success of the proposed algorithm as it outperforms the state-of-the-art methods.

U2 - 10.1109/ICPR.2016.7900200

DO - 10.1109/ICPR.2016.7900200

M3 - Conference contribution

BT - 23rd International Conference on Pattern Recognition (ICPR 2016)

PB - IEEE

T2 - International Conference on Pattern Recognition

Y2 - 1 January 1900

ER -

Joint K-Means Quantization for Approximate Nearest Neighbor Search

Abstract

Conference

Publication forum classification

Access to Document

Fingerprint

Cite this