Improved LSH Algorithm for Approximate Nearest Neighbor Search of High Dimensional Vectors
Project/Area Number |
23680008
|
Research Category |
Grant-in-Aid for Young Scientists (A)
|
Allocation Type | Single-year Grants |
Research Field |
Media informatics/Database
|
Research Institution | Future University-Hakodate |
Principal Investigator |
TERASAWA Kengo 公立はこだて未来大学, システム情報科学部, 准教授 (10435985)
|
Project Period (FY) |
2011-04-01 – 2015-03-31
|
Project Status |
Completed (Fiscal Year 2014)
|
Budget Amount *help |
¥19,630,000 (Direct Cost: ¥15,100,000、Indirect Cost: ¥4,530,000)
Fiscal Year 2014: ¥4,940,000 (Direct Cost: ¥3,800,000、Indirect Cost: ¥1,140,000)
Fiscal Year 2013: ¥5,200,000 (Direct Cost: ¥4,000,000、Indirect Cost: ¥1,200,000)
Fiscal Year 2012: ¥6,630,000 (Direct Cost: ¥5,100,000、Indirect Cost: ¥1,530,000)
Fiscal Year 2011: ¥2,860,000 (Direct Cost: ¥2,200,000、Indirect Cost: ¥660,000)
|
Keywords | アルゴリズム / 画像、文章、音声等認識 / コンテンツ・アーカイブ |
Outline of Final Research Achievements |
In this study, we aim to establish an algorithm to find the nearest-neighbor of the query vector among the huge database of high-dimensional vectors. We applied the SLSH (Spherical LSH) based method for the case of Euclid distance and the improved LSH based method for the intersection similarity, and confirmed that our method is efficient than existing methods.
|
Report
(5 results)
Research Products
(3 results)