English   español  
Please use this identifier to cite or link to this item: http://hdl.handle.net/10261/195639
logo share SHARE logo core CORE   Add this article to your Mendeley library MendeleyBASE

Visualizar otros formatos: MARC | Dublin Core | RDF | ORE | MODS | METS | DIDL
Exportar a otros formatos:


Optimum selection of DNN model and framework for edge inference

AuthorsVelasco-Montero, Delia; Fernández-Berni, J. ; Carmona-Galán, R. ; Rodríguez-Vázquez, Ángel
Keywordsedge inference
convolutional neural networks
embedded vision
deep learning
high-level specifications
Issue Date24-Sep-2018
PublisherInstitute of Electrical and Electronics Engineers
CitationIEEE Access 6: 51680-51692 (2018)
AbstractThis paper describes a methodology to select the optimum combination of deep neural network and software framework for visual inference on embedded systems. As a first step, benchmarking is required. In particular, we have benchmarked six popular network models running on four deep learning frameworks implemented on a low-cost embedded platform. Three key performance metrics have been measured and compared with the resulting 24 combinations: accuracy, throughput, and power consumption. Then, application-level specifications come into play. We propose a figure of merit enabling the evaluation of each network/framework pair in terms of relative importance of the aforementioned metrics for a targeted application. We prove through numerical analysis and meaningful graphical representations that only a reduced subset of the combinations must actually be considered for real deployment. Our approach can be extended to other networks, frameworks, and performance parameters, thus supporting system-level design decisions in the ever-changing ecosystem of embedded deep learning technology.
Publisher version (URL)https://doi.org/10.1109/ACCESS.2018.2869929
Appears in Collections:(IMSE-CNM) Artículos
Files in This Item:
File Description SizeFormat 
FINAL Article.pdf2,75 MBAdobe PDFThumbnail
Show full item record
Review this work

Related articles:

WARNING: Items in Digital.CSIC are protected by copyright, with all rights reserved, unless otherwise indicated.