EUVIMEDEUVIMED entdecken
ÜbersichtsarbeitOpen Access · CC BY 4.0

Binary Neural Networks in FPGAs: Architectures, Tool Flows and Hardware Comparisons

Yuanxin Su, Kah Phooi Seng, Li Minn Ang, Jeremy Smith

Sensors · 17.11.2023 · 23 (22) · 9254

DOI: 10.3390/s23229254 ↗

Kurz gefasst

Übersichtsarbeit in der Zeitschrift Sensors, online erschienen am 17.11.2023. Verfasst von Yuanxin Su und 3 weiteren Autor:innen. Beteiligte Einrichtung: School of AI and Advanced Computing, Xi'an Jiaotong Liverpool University, Suzhou 215000, China. Die Arbeit wurde bisher 16-mal zitiert (Crossref) und nennt 78 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Binary neural networks (BNNs) are variations of artificial/deep neural network (ANN/DNN) architectures that constrain the real values of weights to the binary set of numbers {−1,1}. By using binary values, BNNs can convert matrix multiplications into bitwise operations, which accelerates both training and inference and reduces hardware complexity and model sizes for implementation. Compared to traditional deep learning architectures, BNNs are a good choice for implementation in resource-constrained devices like FPGAs and ASICs. However, BNNs have the disadvantage of reduced performance and accuracy because of the tradeoff due to binarization. Over the years, this has attracted the attention of the research community to overcome the performance gap of BNNs, and several architectures have been proposed. In this paper, we provide a comprehensive review of BNNs for implementation in FPGA hardware. The survey covers different aspects, such as BNN architectures and variants, design and tool flows for FPGAs, and various applications for BNNs. The final part of the paper gives some benchmark works and design tools for implementing BNNs in FPGAs based on established datasets used by the research community.

Schlagwörter

Stichwörter der Autor:innen

binary neural network (BNN)computational modelingfield-programmable gate array (FPGA)latency reduction

Autor:innen und Institutionen

  • School of AI and Advanced Computing, Xi'an Jiaotong Liverpool University, Suzhou 215000, China.
  • Department of Electrical Engineering and Electronics, University of Liverpool, Liverpool L69 3GJ, UK.
  • School of Computer Science, Queensland University of Technology, Brisbane City, QLD 4000, Australia.
  • School of Science Technology and Engineering, University of the Sunshine Coast, Sippy Downs, QLD 4556, Australia.

Zitieren

APA

Su, Y., Seng, K. P., Ang, L. M., & Smith, J. (2023). Binary Neural Networks in FPGAs: Architectures, Tool Flows and Hardware Comparisons. Sensors, 23(22), 9254. https://doi.org/10.3390/s23229254

Vancouver

Su Y, Seng KP, Ang LM, Smith J. Binary Neural Networks in FPGAs: Architectures, Tool Flows and Hardware Comparisons. Sensors. 2023;23(22):9254. doi:10.3390/s23229254

BibTeX
@article{Su2023,
  author = {Su, Yuanxin and Seng, Kah Phooi and Ang, Li Minn and Smith, Jeremy},
  title = {Binary Neural Networks in FPGAs: Architectures, Tool Flows and Hardware Comparisons},
  journal = {Sensors},
  year = {2023},
  volume = {23},
  number = {22},
  pages = {9254},
  doi = {10.3390/s23229254}
}
Kostenlos starten →