Quantization Effects of Deep Neural Networks on a FPGA platform

Daniel Pohl, Birgit Vogel-Heuser, Marius Kruger, Markus Echtler

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, a quantization method for a FPGA platform is applied on three different deep neural networks (DNNs) for classification, detection and semantic segmentation tasks. The FPGA platform is based on a Xilinx Zynq UltraScale+ ZU5EV. The quantization process is done with the Vitis AI Quantizer tool of the AMD Vitis AI Environment. After a short description of the used DNNs, datasets and the technique behind the quantization process, the results are presented for different bit-widths of the quantized DNNs. The evaluation includes a post-training quantization of weights and activations in INT8 format with performance measurements in frames per second (FPS) on the target platform.

Original languageEnglish
Title of host publication2024 IEEE 7th International Conference on Industrial Cyber-Physical Systems, ICPS 2024
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798350363012
DOIs
StatePublished - 2024
Event7th IEEE International Conference on Industrial Cyber-Physical Systems, ICPS 2024 - St. Louis, United States
Duration: 12 May 202415 May 2024

Publication series

Name2024 IEEE 7th International Conference on Industrial Cyber-Physical Systems, ICPS 2024

Conference

Conference7th IEEE International Conference on Industrial Cyber-Physical Systems, ICPS 2024
Country/TerritoryUnited States
CitySt. Louis
Period12/05/2415/05/24

Keywords

  • classification
  • deep learning
  • detection
  • FPGA
  • neural networks
  • quantization
  • segmentation

Fingerprint

Dive into the research topics of 'Quantization Effects of Deep Neural Networks on a FPGA platform'. Together they form a unique fingerprint.

Cite this