Training Large-Scale Optical Neural Networks with Two-Pass Forward Propagation

Amirreza Ahmadnejad, Somayyeh Koohi
8/15/2024
cs.LGphysics.optics

No official code found. Best implementation: AAhmadnejad98/Training-Large-Scale-ONN (1 star, last commit 2024-08-11).

Abstract

This paper addresses the limitations in Optical Neural Networks (ONNs) related to training efficiency, nonlinear function implementation, and large input data processing. We introduce Two-Pass Forward Propagation, a novel training method that avoids specific nonlinear activation functions by modulating and re-entering error with random noise. Additionally, we propose a new way to implement convolutional neural networks using simple neural networks in integrated optical systems. Theoretical foundations and numerical results demonstrate significant improvements in training speed, energy efficiency, and scalability, advancing the potential of optical computing for complex data tasks.

View on arXivView PDF

Code Implementations(1)

AAhmadnejad98/Training-Large-Scale-ONNCommunity100%

Training Large-Scale Optical Neural Networks with Two-Pass Forward Propagation

11MATLAB, Jupyter NotebookAug 2, 20242 years ago

Similar Papers

Forward-Forward Training of an Optical Neural Network

89% match2023

Universality of physical neural networks with multivariate nonlinearity

83% match2025

Backpropagation-free Training of Deep Physical Neural Networks

83% match2023

Artificial Neural Networks for Photonic Applications: From Algorithms to Implementation

83% match2024

Streamlined optical training of large-scale modern deep learning architectures with direct feedback alignment

83% match2024

Neural Tangent Knowledge Distillation for Optical Convolutional Networks

83% match2025

Cite this paper

@article{ahmadnejad2024training,
  title  = {Training Large-Scale Optical Neural Networks with Two-Pass Forward Propagation},
  author = {Amirreza Ahmadnejad and Somayyeh Koohi},
  year   = {2024},
  eprint = {2408.08337},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2408.08337v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2408.08337.svg)](https://codeofpaper.com/paper/2408.08337)

Discussion