Tunable Nonlinear Activation Functions for Optical Neural Networks

Ian A.D. Williamson; Tyler W. Hughes; Momchil Minkov; Ben Bartlett; Sunil Pai; Shanhui Fan

Tunable Nonlinear Activation Functions for Optical Neural Networks

Ian A.D. Williamson, Tyler W. Hughes, Momchil Minkov, Ben Bartlett, Sunil Pai, Shanhui Fan

Stanford University

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

Abstract

We introduce an electro-optic hardware platform for realizing optical nonlinearities and demonstrate that, as a nonlinear activation function, it can substantially improve the classification performance of optical neural networks.

Original language	English
Title of host publication	2020 Conference on Lasers and Electro-Optics, CLEO 2020 - Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781943580767
Publication status	Published - May 2020
Externally published	Yes
Event	2020 Conference on Lasers and Electro-Optics, CLEO 2020 - San Jose, United States Duration: 10 May 2020 → 15 May 2020

Publication series

Name	Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS
Volume	2020-May
ISSN (Print)	1092-8081

Conference

Conference	2020 Conference on Lasers and Electro-Optics, CLEO 2020
Country/Territory	United States
City	San Jose
Period	10/05/20 → 15/05/20

Cite this

Williamson, I. A. D., Hughes, T. W., Minkov, M., Bartlett, B., Pai, S., & Fan, S. (2020). Tunable Nonlinear Activation Functions for Optical Neural Networks. In 2020 Conference on Lasers and Electro-Optics, CLEO 2020 - Proceedings Article 9192330 (Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS; Vol. 2020-May). Institute of Electrical and Electronics Engineers Inc..

@inproceedings{74992a62bed64ec28c4c14c4fc95ffc0,

title = "Tunable Nonlinear Activation Functions for Optical Neural Networks",

abstract = "We introduce an electro-optic hardware platform for realizing optical nonlinearities and demonstrate that, as a nonlinear activation function, it can substantially improve the classification performance of optical neural networks.",

author = "Williamson, {Ian A.D.} and Hughes, {Tyler W.} and Momchil Minkov and Ben Bartlett and Sunil Pai and Shanhui Fan",

note = "Publisher Copyright: {\textcopyright} 2020 OSA.; 2020 Conference on Lasers and Electro-Optics, CLEO 2020 ; Conference date: 10-05-2020 Through 15-05-2020",

year = "2020",

month = may,

language = "English",

series = "Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2020 Conference on Lasers and Electro-Optics, CLEO 2020 - Proceedings",

}

Williamson, IAD, Hughes, TW, Minkov, M, Bartlett, B, Pai, S & Fan, S 2020, Tunable Nonlinear Activation Functions for Optical Neural Networks. in 2020 Conference on Lasers and Electro-Optics, CLEO 2020 - Proceedings., 9192330, Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS, vol. 2020-May, Institute of Electrical and Electronics Engineers Inc., 2020 Conference on Lasers and Electro-Optics, CLEO 2020, San Jose, United States, 10/05/20.

Tunable Nonlinear Activation Functions for Optical Neural Networks. / Williamson, Ian A.D.; Hughes, Tyler W.; Minkov, Momchil et al.
2020 Conference on Lasers and Electro-Optics, CLEO 2020 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2020. 9192330 (Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS; Vol. 2020-May).

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

TY - GEN

T1 - Tunable Nonlinear Activation Functions for Optical Neural Networks

AU - Williamson, Ian A.D.

AU - Hughes, Tyler W.

AU - Minkov, Momchil

AU - Bartlett, Ben

AU - Pai, Sunil

AU - Fan, Shanhui

PY - 2020/5

Y1 - 2020/5

N2 - We introduce an electro-optic hardware platform for realizing optical nonlinearities and demonstrate that, as a nonlinear activation function, it can substantially improve the classification performance of optical neural networks.

AB - We introduce an electro-optic hardware platform for realizing optical nonlinearities and demonstrate that, as a nonlinear activation function, it can substantially improve the classification performance of optical neural networks.

UR - http://www.scopus.com/inward/record.url?scp=85091647476&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85091647476

T3 - Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS

BT - 2020 Conference on Lasers and Electro-Optics, CLEO 2020 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2020 Conference on Lasers and Electro-Optics, CLEO 2020

Y2 - 10 May 2020 through 15 May 2020

ER -

Tunable Nonlinear Activation Functions for Optical Neural Networks

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this