Ciliberto, Mathias and Roggen, Daniel (2019) WLCSSCuda: a CUDA accelerated template matching method for gesture recognition. ISWC 2019: International Symposium on Wearable Computers, London, 9-13 September 2019. Published in: Dunne, Lucy, Cakmakci, Ozan, Ward, Jamie, Unset, Unset, Unset, Unset and Unset, (eds.) Proceedings of the 2019 International Symposium on Wearable Computers. 32-34. Association for Computing Machinery ISBN 9781450368704

PDF - Accepted Version Available under License All Rights Reserved. Download (558kB)

Abstract

Template matching methods can benefit from multi-cores architecture in order to parallelise and accelerate the matching of multiple templates. We present WLCSSCuda: a GPU accelerated implementation of the Warping Longest Common Subsequence (WLCSS) pattern recognition algorithm. We evaluate our method on 4 NVIDIA GPUs and 4 multi-cores CPUs. We observe a 67-times speedup for the GPU implementation in the best case against the multithreaded CPU implementation.

Item Type:	Conference Proceedings
Keywords:	WLCSS; CUDA; GPU acceleration; Template Matching
Schools and Departments:	School of Engineering and Informatics > Engineering and Design
Research Centres and Groups:	Sensor Technology Research Centre
Subjects:	Q Science > QA Mathematics T Technology > T Technology (General)
Related URLs:	Organisation
Depositing User:	Daniel Roggen
Date Deposited:	08 Aug 2019 09:49
Last Modified:	21 Nov 2019 15:52
URI:	http://sro.sussex.ac.uk/id/eprint/85365

View download statistics for this item

📧 Request an update