loading page

Photonic convolutional neural networks using integrated diffractive optics
  • +2
  • Jun Rong Ong ,
  • Thomas Yong Long Ang ,
  • Chin Chun Ooi ,
  • Soon Thor Lim ,
  • Ching Eng Png
Jun Rong Ong
Institute of High Performance Computing

Corresponding Author:[email protected]

Author Profile
Thomas Yong Long Ang
Author Profile
Chin Chun Ooi
Author Profile
Soon Thor Lim
Author Profile
Ching Eng Png
Author Profile

Abstract

With recent rapid advances in photonic integrated circuits, it has been demonstrated that programmable photonic chips can be used to implement artificial neural networks. Convolutional neural networks (CNN) are a class of deep learning methods that have been highly successful in applications such as image classification and speech processing. We present an architecture to implement a photonic CNN using the Fourier transform property of integrated star couplers. We show, in computer simulation, high accuracy image classification using the MNIST dataset. We also model component imperfections in photonic CNN and show that the performance degradation can be recovered in a programmable chip. Our proposed architecture provides a large reduction in physical footprint compared to current implementations as it utilizes the natural advantages of optics and hence offers a scalable pathway towards integrated photonic deep learning processors.
Sep 2020Published in IEEE Journal of Selected Topics in Quantum Electronics volume 26 issue 5 on pages 1-8. 10.1109/JSTQE.2020.2982990