Range query processing on single and multi GPU environments

Barrientos, Ricardo J.; Gómez, José Ignacio Aguaded; Tenllado, Christian; Prieto-Matías, Manuel

doi:10.1016/j.compeleceng.2013.05.012

Cited by 6 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are several parallel algorithms that use the LC index on different kinds of parallel platforms: distributed memory [11] and shared memory [9] multiprocessors, GPUs [6] and multi-GPU platforms [7]. We will compare our parallel algorithms with the local version of the algorithm introduced in [9] to deal with high query traffic.…”

Section: Related Workmentioning

confidence: 99%

Accessing very high dimensional spaces in parallel

Artigas-Fuentes

Badía

2016

J Supercomput

View full text Add to dashboard Cite

Access methods are a fundamental tool on Information Retrieval. However, most of these methods suffer the problem known as the curse of dimensionality when they are applied to objects with very high dimensionality representation spaces, such as text documents. In this paper we introduce a new parallel access method that uses several graphs as distributed index structure and a kN N search algorithm. Two parallel versions of the search method are presented, one based on master-slave scheme and the other based on a pipeline. A thorough experimental analysis on different datasets shows that our method can process efficiently large flows of queries, compete with other parallel algorithms and obtain at the same time very high quality results.

show abstract

Section: Related Workmentioning

confidence: 99%

Accessing very high dimensional spaces in parallel

Artigas-Fuentes

Badía

2016

J Supercomput

View full text Add to dashboard Cite

show abstract

“…Thus, it implies 45 seconds for a database with 1000 fingerprints, which is a considerable time, especially when the database reaches the order of tens thousands or more fingerprints. There are two usual methods to decrease the execution time in this case, which are (1) to avoid comparing all fingerprint pairs by using classification methods [8,16] or indexing algorithms [7,4], and (2) to accelerate the matching processing by using parallel computing [27,21].…”

Section: Introductionmentioning

confidence: 99%

Heap-Based Algorithms to Accelerate Fingerprint Matching on Parallel Platforms

Barrientos

Hernández-García

Ortega

et al. 2019

Communications in Computer and Information Science

Self Cite

View full text Add to dashboard Cite

Nowadays, fingerprint is the most used biometric trait for in dividuals identification. In this area, the state-of-the-art algorithms are very accurate, but when the database contains millions of identities, an acceleration of the algorithm is required. From these algorithms, Minutia Cylinder-Code (MCC) stands out for its good results in terms of accu racy, however its efficiency in computational time is not high. In this work, we propose to use two different parallel platforms to accelerate fingerprint matching process by using MCC: (1) a multi-core server, and (2) a Xeon Phi coprocessor. Our proposal is based on heaps as auxiliary structure to process the global similarity of MCC. As heap-based algo rithms are exhaustive (all the elements are accessed), we also explored the use an indexing algorithm to avoid comparing the query against all the fingerprints of the database. Experimental results show an improve ment up to 97.15x of speed-up, which is competitive compared to other state-of-the-art algorithms in GPU and FPGA. To the best of our knowl edge, this is the first work for fingerprint identification using a Xeon Phi coprocessor.

show abstract