Saved in:
Bibliographic Details
Main Authors: Pes, Lorenzo, Chehreghan, Maryam Dehbashizadeh, Luiken, Rick, Stuijk, Sander, Stabile, Ripalta, Corradi, Federico
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.04141
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909451047403520
author Pes, Lorenzo
Chehreghan, Maryam Dehbashizadeh
Luiken, Rick
Stuijk, Sander
Stabile, Ripalta
Corradi, Federico
author_facet Pes, Lorenzo
Chehreghan, Maryam Dehbashizadeh
Luiken, Rick
Stuijk, Sander
Stabile, Ripalta
Corradi, Federico
contents This work evaluates a forward-only learning algorithm on the MNIST dataset with hardware-in-the-loop training of a 4f optical correlator, achieving 87.6% accuracy with O(n2) complexity, compared to backpropagation, which achieves 88.8% accuracy with O(n2 log n) complexity.
format Preprint
id arxiv_https___arxiv_org_abs_2501_04141
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Hardware-In-The-Loop Training of a 4f Optical Correlator with Logarithmic Complexity Reduction for CNNs
Pes, Lorenzo
Chehreghan, Maryam Dehbashizadeh
Luiken, Rick
Stuijk, Sander
Stabile, Ripalta
Corradi, Federico
Neural and Evolutionary Computing
This work evaluates a forward-only learning algorithm on the MNIST dataset with hardware-in-the-loop training of a 4f optical correlator, achieving 87.6% accuracy with O(n2) complexity, compared to backpropagation, which achieves 88.8% accuracy with O(n2 log n) complexity.
title Hardware-In-The-Loop Training of a 4f Optical Correlator with Logarithmic Complexity Reduction for CNNs
topic Neural and Evolutionary Computing
url https://arxiv.org/abs/2501.04141