Presentation 2013-11-13
Gaussian Sparse Hashing
Koichiro SUZUKI, Mitsuru AMBAI, Ikuro SATO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Binary hashing has been widely used for Approximate Nearest Neighbor (ANN) search due to fast query speed and less storage cost. It is important for hashing functions to preserve a similarity of data space. We introduce a novel concept for preserving the similarity, a robustness index and an error cost function, and propose a hashing function learning method by optimizing the cost function with an assumption the data has Gaussian probability density function. Furthermore, we expand this method to learn a sparse hash function for faster coding. It is demonstrated that proposed methods have comparable level of similarity preservation to the existing methods.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) binary codes / approximate image retrieval / approximate nearest neighbor searching / sparse matrix
Paper # IBISML2013-56
Date of Issue

Conference Information
Committee IBISML
Conference Date 2013/11/5(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Information-Based Induction Sciences and Machine Learning (IBISML)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Gaussian Sparse Hashing
Sub Title (in English)
Keyword(1) binary codes
Keyword(2) approximate image retrieval
Keyword(3) approximate nearest neighbor searching
Keyword(4) sparse matrix
1st Author's Name Koichiro SUZUKI
1st Author's Affiliation DENSO ITLABORATORY, INC.()
2nd Author's Name Mitsuru AMBAI
2nd Author's Affiliation DENSO ITLABORATORY, INC.
3rd Author's Name Ikuro SATO
3rd Author's Affiliation DENSO ITLABORATORY, INC.
Date 2013-11-13
Paper # IBISML2013-56
Volume (vol) vol.113
Number (no) 286
Page pp.pp.-
#Pages 6
Date of Issue