Dual-Precision Deep Neural Network

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

1 Scopus citations

Abstract

On-line Precision scalability of the deep neural networks(DNNs) is a critical feature to support accuracy and complexity trade-off during the DNN inference. In this paper, we propose dual-precision DNN that includes two different precision modes in a single model, thereby supporting an on-line precision switch without re-training. The proposed two-phase training process optimizes both low- and high-precision modes.

Original languageEnglish
Title of host publicationProceedings of the 2020 3rd International Conference on Artificial Intelligence and Pattern Recognition, AIPR 2020
PublisherAssociation for Computing Machinery
Pages30-34
Number of pages5
ISBN (Electronic)9781450375511
DOIs
StatePublished - 26 Jun 2020
Event3rd International Conference on Artificial Intelligence and Pattern Recognition, AIPR 2020 - Virtual, Online, China
Duration: 26 Jun 202028 Jun 2020

Publication series

NameACM International Conference Proceeding Series

Conference

Conference3rd International Conference on Artificial Intelligence and Pattern Recognition, AIPR 2020
Country/TerritoryChina
CityVirtual, Online
Period26/06/2028/06/20

Keywords

  • Deep neural network
  • dual precision
  • precision scalable
  • weight quantization

Fingerprint

Dive into the research topics of 'Dual-Precision Deep Neural Network'. Together they form a unique fingerprint.

Cite this