An Optical Computing Chip for Executing Complex-valued Neural Network and Its On-chip Training

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review

Abstract

The optical implementation of neural networks is proposed to have advantages over electronic implementations with lower power consumption and higher computation speed. However, most optical neural networks (ONNs) utilize conventional real-valued frameworks that are designed for digital computers, forfeiting many advantages of optical computing such as efficient complex-valued operations. Complex-valued neural networks are advantageous to their real-valued counterparts by offering rich representation space, fast convergence, and strong generalizations. We propose and demonstrate an ONN that implements truly complex-valued neural networks, achieving high accuracy and strong learning capability in many benchmark tasks [1]. On the other hand, efficiently training ONNs remains a formidable challenge, due to the difficulty in obtaining gradient information from a physical device. We propose an efficient on-chip training protocol for ONNs and demonstrate it by several practical tasks [2]. The protocol is gradient-free and physical agnostic, and is applicable for various types of chip structures, especially those that cannot be analytically decomposed and characterized. The protocol is robust to experimental perturbations like imperfect phase detection and photodetection noise. Our results present a promising avenue towards deep complex networks with smaller chip size, stronger performance, and flexible reconfiguration to realistic applications (e.g., facial recognition, natural language processing, and autonomous vehicles).

Original languageEnglish
Title of host publication2022 Photonics and Electromagnetics Research Symposium, PIERS 2022 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages189-196
Number of pages8
ISBN (Electronic)9781665460231
DOIs
Publication statusPublished - Apr 2022
Externally publishedYes
Event2022 Photonics and Electromagnetics Research Symposium, PIERS 2022 - Hangzhou, China
Duration: 25 Apr 202229 Apr 2022

Publication series

NameProgress in Electromagnetics Research Symposium
Volume2022-April
ISSN (Print)1559-9450
ISSN (Electronic)1931-7360

Conference

Conference2022 Photonics and Electromagnetics Research Symposium, PIERS 2022
Country/TerritoryChina
CityHangzhou
Period25/04/2229/04/22

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Electronic, Optical and Magnetic Materials

Fingerprint

Dive into the research topics of 'An Optical Computing Chip for Executing Complex-valued Neural Network and Its On-chip Training'. Together they form a unique fingerprint.

Cite this