Joint Optimization of Dimension Reduction and Mixed-Precision Quantization for Activation Compression of Neural Networks-Reference-Cited by-同舟云学术

Joint Optimization of Dimension Reduction and Mixed-Precision Quantization for Activation Compression of Neural Networks

Published:2023-11 Issue:11 Volume:42 Page:4025-4037
ISSN:0278-0070
Container-title:IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
language:
Short-container-title:IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.

Author:

Tai Yu-Shan¹^ORCID,Chang Cheng-Yang¹^ORCID,Teng Chieh-Fang²^ORCID,Chen Yi-Ta¹^ORCID,Wu An-Yeu¹^ORCID

Affiliation:

1. GIEE, National Taiwan University, Taipei, Taiwan

2. CAI2, Mediatek, Hsinchu, Taiwan

Funder

Novatek

Ministry of Science and Technology, Taiwan

NSTC

Synopsys Ph.D. Talent Cultivation Scholarship

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Subject

Electrical and Electronic Engineering,Computer Graphics and Computer-Aided Design,Software

Link

Reference28 articles.

2. DoReFa-net: Training low bitwidth convolutional neural networks with low bitwidth gradients;zhou;arXiv 1606 06160 [cs],2016

4. Binarized neural networks;hubara;Proc Int Conf Neural Inf Process Syst (NeurIPS),2016

Cited by 2 articles. 订阅此论文施引文献订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献