EnglishisÍslenska

Member institutions

Search in


ReportReykjavík University>Tölvunarfræðideild>Reykjavik University Technical Report Computer Science. RUTR-CS>

Please use this identifier to cite or link to this item: http://hdl.handle.net/1946/7496

Title

Full GPU acceleration of local descriptors using CUDA

Published
July 2009
Abstract

Video analysis using local decriptors demands an high-throughput descirptor creation process. The most practical methods to this goal is to use GPUs that come with most resent computers. In this report, we adapt the computation of the Eff2 descriptors, a variant of SIFT, to the GPU. We compare our GPU-Rff2 descriptors to SiftGPU and show that while both GPU-based variants yield similar results, the GPU-Eff2 descriptors require sifnificantly less processing time.

Appeared in

RUTR-CS09003

ISSN

1670-5777

Issued Date
01/02/2011


Artifacts
Name[Sortable]Size[Sortable]Visibility[Sortable]Description[Sortable]Format
RUTR-CS09003.pdf2.11MBOpen Complete Text PDF View/Open