On Building Efficient and Robust Neural Network Designs

Xiaoxuan Yang; Huanrui Yang; Jingchi Zhang; Hai Helen Li; Yiran Chen

doi:10.1109/IEEECONF56349.2022.10051891

On Building Efficient and Robust Neural Network Designs

Xiaoxuan Yang
, Huanrui Yang
, Jingchi Zhang
, Hai Helen Li
, Yiran Chen

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

2 Scopus citations

Abstract

Neural network models have demonstrated outstanding performance in a variety of applications, from image classification to natural language processing. However, deploying the models to hardware raises efficiency and reliability issues. From the efficiency perspective, the storage, computation, and communication cost of neural network processing is considerably large because the neural network models have a large number of parameters and operations. From the standpoint of robustness, the perturbation in hardware is unavoidable and thus the performance of neural networks can be degraded. As a result, this paper investigates effective learning and optimization approaches as well as advanced hardware designs in order to build efficient and robust neural network designs.

Original language	English (US)
Title of host publication	56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022
Editors	Michael B. Matthews
Publisher	IEEE Computer Society
Pages	317-321
Number of pages	5
ISBN (Electronic)	9781665459068
DOIs	https://doi.org/10.1109/IEEECONF56349.2022.10051891
State	Published - 2022
Externally published	Yes
Event	56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022 - Virtual, Online, United States Duration: Oct 31 2022 → Nov 2 2022

Publication series

Name	Conference Record - Asilomar Conference on Signals, Systems and Computers
Volume	2022-October
ISSN (Print)	1058-6393

Conference

Conference	56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022
Country/Territory	United States
City	Virtual, Online
Period	10/31/22 → 11/2/22

Keywords

efficiency
hardware-software co-design
neural network
robustness

ASJC Scopus subject areas

Signal Processing
Computer Networks and Communications

Access to Document

10.1109/IEEECONF56349.2022.10051891

Cite this

Yang, X., Yang, H., Zhang, J., Li, H. H., & Chen, Y. (2022). On Building Efficient and Robust Neural Network Designs. In M. B. Matthews (Ed.), 56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022 (pp. 317-321). (Conference Record - Asilomar Conference on Signals, Systems and Computers; Vol. 2022-October). IEEE Computer Society. https://doi.org/10.1109/IEEECONF56349.2022.10051891

On Building Efficient and Robust Neural Network Designs. / Yang, Xiaoxuan; Yang, Huanrui; Zhang, Jingchi et al.
56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022. ed. / Michael B. Matthews. IEEE Computer Society, 2022. p. 317-321 (Conference Record - Asilomar Conference on Signals, Systems and Computers; Vol. 2022-October).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Yang, X, Yang, H, Zhang, J, Li, HH & Chen, Y 2022, On Building Efficient and Robust Neural Network Designs. in MB Matthews (ed.), 56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022. Conference Record - Asilomar Conference on Signals, Systems and Computers, vol. 2022-October, IEEE Computer Society, pp. 317-321, 56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022, Virtual, Online, United States, 10/31/22. https://doi.org/10.1109/IEEECONF56349.2022.10051891

@inproceedings{94fca71220b7427f861cb267989a4fa4,

title = "On Building Efficient and Robust Neural Network Designs",

abstract = "Neural network models have demonstrated outstanding performance in a variety of applications, from image classification to natural language processing. However, deploying the models to hardware raises efficiency and reliability issues. From the efficiency perspective, the storage, computation, and communication cost of neural network processing is considerably large because the neural network models have a large number of parameters and operations. From the standpoint of robustness, the perturbation in hardware is unavoidable and thus the performance of neural networks can be degraded. As a result, this paper investigates effective learning and optimization approaches as well as advanced hardware designs in order to build efficient and robust neural network designs.",

keywords = "efficiency, hardware-software co-design, neural network, robustness",

author = "Xiaoxuan Yang and Huanrui Yang and Jingchi Zhang and Li, \{Hai Helen\} and Yiran Chen",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022 ; Conference date: 31-10-2022 Through 02-11-2022",

year = "2022",

doi = "10.1109/IEEECONF56349.2022.10051891",

language = "English (US)",

series = "Conference Record - Asilomar Conference on Signals, Systems and Computers",

publisher = "IEEE Computer Society",

pages = "317--321",

editor = "Matthews, \{Michael B.\}",

booktitle = "56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022",

}

TY - GEN

T1 - On Building Efficient and Robust Neural Network Designs

AU - Yang, Xiaoxuan

AU - Yang, Huanrui

AU - Zhang, Jingchi

AU - Li, Hai Helen

AU - Chen, Yiran

PY - 2022

Y1 - 2022

N2 - Neural network models have demonstrated outstanding performance in a variety of applications, from image classification to natural language processing. However, deploying the models to hardware raises efficiency and reliability issues. From the efficiency perspective, the storage, computation, and communication cost of neural network processing is considerably large because the neural network models have a large number of parameters and operations. From the standpoint of robustness, the perturbation in hardware is unavoidable and thus the performance of neural networks can be degraded. As a result, this paper investigates effective learning and optimization approaches as well as advanced hardware designs in order to build efficient and robust neural network designs.

AB - Neural network models have demonstrated outstanding performance in a variety of applications, from image classification to natural language processing. However, deploying the models to hardware raises efficiency and reliability issues. From the efficiency perspective, the storage, computation, and communication cost of neural network processing is considerably large because the neural network models have a large number of parameters and operations. From the standpoint of robustness, the perturbation in hardware is unavoidable and thus the performance of neural networks can be degraded. As a result, this paper investigates effective learning and optimization approaches as well as advanced hardware designs in order to build efficient and robust neural network designs.

KW - efficiency

KW - hardware-software co-design

KW - neural network

KW - robustness

UR - https://www.scopus.com/pages/publications/85150192003

UR - https://www.scopus.com/pages/publications/85150192003#tab=citedBy

U2 - 10.1109/IEEECONF56349.2022.10051891

DO - 10.1109/IEEECONF56349.2022.10051891

M3 - Conference contribution

AN - SCOPUS:85150192003

T3 - Conference Record - Asilomar Conference on Signals, Systems and Computers

SP - 317

EP - 321

BT - 56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022

A2 - Matthews, Michael B.

PB - IEEE Computer Society

T2 - 56th Asilomar Conference on Signals, Systems and Computers, ACSSC 2022

Y2 - 31 October 2022 through 2 November 2022

ER -

On Building Efficient and Robust Neural Network Designs

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this