Publication detail

Comparison of FPNNs Models Approximation Capabilities and FPGA Resources Utilization

KRČMA, M. KOTÁSEK, Z. LOJDA, J.

Original Title

Comparison of FPNNs Models Approximation Capabilities and FPGA Resources Utilization

Type

conference paper

Language

English

Original Abstract

This paper presents the concepts of FPNA and FPNN, used for the approximation of artificial neural networks in FPGAs and introduces derived types of these concepts used by the authors. The process of transformation of a trained artificial neural network to an FPNN is  described. The diagram of the FPGA implementation is presented. The results of experiments determining the approximation capabilities of FPNNs are presented and the FPGA resources utilization are compared.

Keywords

ANN, FPNN, FPGA

Authors

KRČMA, M.; KOTÁSEK, Z.; LOJDA, J.

Released

6. 9. 2017

Publisher

IEEE Computer Society

Location

Cluj-Nappoca

ISBN

978-1-5386-3368-7

Book

Proceedings of IEEE 13th International Conference on Intelligent Computer Communication and Processing

Pages from

125

Pages to

132

Pages count

8

URL

BibTex

@inproceedings{BUT146266,
  author="Martin {Krčma} and Zdeněk {Kotásek} and Jakub {Lojda}",
  title="Comparison of FPNNs Models Approximation Capabilities and FPGA Resources Utilization",
  booktitle="Proceedings of  IEEE 13th International Conference on Intelligent Computer Communication and Processing",
  year="2017",
  pages="125--132",
  publisher="IEEE Computer Society",
  address="Cluj-Nappoca",
  doi="10.1109/ICCP.2017.8116993",
  isbn="978-1-5386-3368-7",
  url="https://www.fit.vut.cz/research/publication/11507/"
}