Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

LightFF: Lightweight Inference for Forward-Forward Algorithm

Aminifar, Amin ; Huang, Baichuan LU orcid ; Abtahi Fahliani, Azra LU and Aminifar, Amir LU orcid (2024) 27th European Conference on Artificial Intelligence, ECAI-2024 In Frontiers in Artificial Intelligence and Applications 392. p.1728-1735
Abstract
The human brain performs tasks with an outstanding energy efficiency, i.e., with approximately 20 Watts. The state-of-the-art Artificial/Deep Neural Networks (ANN/DNN), on the other hand, have recently been shown to consume massive amounts of energy. The training of these ANNs/DNNs is done almost exclusively based on the back-propagation algorithm, which is known to be biologically implausible. This has led to a new generation of forward-only techniques, including the Forward-Forward algorithm. In this paper, we propose a lightweight inference scheme specifically designed for DNNs trained using the Forward-Forward algorithm. We have evaluated our proposed lightweight inference scheme in the case of the MNIST and CIFAR datasets, as well as... (More)
The human brain performs tasks with an outstanding energy efficiency, i.e., with approximately 20 Watts. The state-of-the-art Artificial/Deep Neural Networks (ANN/DNN), on the other hand, have recently been shown to consume massive amounts of energy. The training of these ANNs/DNNs is done almost exclusively based on the back-propagation algorithm, which is known to be biologically implausible. This has led to a new generation of forward-only techniques, including the Forward-Forward algorithm. In this paper, we propose a lightweight inference scheme specifically designed for DNNs trained using the Forward-Forward algorithm. We have evaluated our proposed lightweight inference scheme in the case of the MNIST and CIFAR datasets, as well as two real-world applications, namely, epileptic seizure detection and cardiac arrhythmia classification using wearable technologies, where complexity overheads/energy consumption is a major constraint, and demonstrate its relevance. Our code is available at this https URL. https://github.com/AminAminifar/LightFF (Less)
Please use this url to cite or link to this publication:
author
; ; and
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
host publication
27th European Conference on Artificial Intelligence, 19–24 October 2024, Santiago de Compostela, Spain – Including 13th Conference on Prestigious Applications of Intelligent Systems (PAIS 2024)
series title
Frontiers in Artificial Intelligence and Applications
volume
392
pages
1728 - 1735
publisher
IOS Press
conference name
27th European Conference on Artificial Intelligence, ECAI-2024
conference location
Santiago de Campostela, Spain
conference dates
2024-10-19 - 2024-10-24
external identifiers
  • scopus:85213328802
ISSN
1879-8314
0922-6389
ISBN
978-1-64368-548-9
DOI
10.3233/FAIA240682
language
English
LU publication?
yes
id
08dbbe15-482b-4747-80f9-b649395236b7
date added to LUP
2024-08-25 09:03:50
date last changed
2025-06-13 16:20:58
@inproceedings{08dbbe15-482b-4747-80f9-b649395236b7,
  abstract     = {{The human brain performs tasks with an outstanding energy efficiency, i.e., with approximately 20 Watts. The state-of-the-art Artificial/Deep Neural Networks (ANN/DNN), on the other hand, have recently been shown to consume massive amounts of energy. The training of these ANNs/DNNs is done almost exclusively based on the back-propagation algorithm, which is known to be biologically implausible. This has led to a new generation of forward-only techniques, including the Forward-Forward algorithm. In this paper, we propose a lightweight inference scheme specifically designed for DNNs trained using the Forward-Forward algorithm. We have evaluated our proposed lightweight inference scheme in the case of the MNIST and CIFAR datasets, as well as two real-world applications, namely, epileptic seizure detection and cardiac arrhythmia classification using wearable technologies, where complexity overheads/energy consumption is a major constraint, and demonstrate its relevance. Our code is available at this https URL. https://github.com/AminAminifar/LightFF}},
  author       = {{Aminifar, Amin and Huang, Baichuan and Abtahi Fahliani, Azra and Aminifar, Amir}},
  booktitle    = {{27th European Conference on Artificial Intelligence, 19–24 October 2024, Santiago de Compostela, Spain – Including 13th Conference on Prestigious Applications of Intelligent Systems (PAIS 2024)}},
  isbn         = {{978-1-64368-548-9}},
  issn         = {{1879-8314}},
  language     = {{eng}},
  pages        = {{1728--1735}},
  publisher    = {{IOS Press}},
  series       = {{Frontiers in Artificial Intelligence and Applications}},
  title        = {{LightFF: Lightweight Inference for Forward-Forward Algorithm}},
  url          = {{https://lup.lub.lu.se/search/files/193938144/LightFF.pdf}},
  doi          = {{10.3233/FAIA240682}},
  volume       = {{392}},
  year         = {{2024}},
}