Early Exiting Predictive Coding Neural Networks for Edge AI

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zniber, Alaa, Ghogho, Mounir, Karrakchou, Ouassim, Zakroum, Mehdi
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908922027180032
author Zniber, Alaa
Ghogho, Mounir
Karrakchou, Ouassim
Zakroum, Mehdi
author_facet Zniber, Alaa
Ghogho, Mounir
Karrakchou, Ouassim
Zakroum, Mehdi
contents The Internet of Things is transforming various fields, with sensors increasingly embedded in wearables, smart buildings, and connected equipment. While deep learning enables valuable insights from IoT data, conventional models are too computationally demanding for resource-limited edge devices. Moreover, privacy concerns and real-time processing needs make local computation a necessity over cloud-based solutions. Inspired by the brain's energy efficiency, we propose a shallow bidirectional predictive coding network with early exiting, dynamically halting computations once a performance threshold is met. This reduces the memory footprint and computational overhead while maintaining high accuracy. We validate our approach using the CIFAR-10 dataset. Our model achieves performance comparable to deep networks with significantly fewer parameters and lower computational complexity, demonstrating the potential of biologically inspired architectures for efficient edge AI.
format Preprint
id arxiv_https___arxiv_org_abs_2309_02022
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Early Exiting Predictive Coding Neural Networks for Edge AI
Zniber, Alaa
Ghogho, Mounir
Karrakchou, Ouassim
Zakroum, Mehdi
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
The Internet of Things is transforming various fields, with sensors increasingly embedded in wearables, smart buildings, and connected equipment. While deep learning enables valuable insights from IoT data, conventional models are too computationally demanding for resource-limited edge devices. Moreover, privacy concerns and real-time processing needs make local computation a necessity over cloud-based solutions. Inspired by the brain's energy efficiency, we propose a shallow bidirectional predictive coding network with early exiting, dynamically halting computations once a performance threshold is met. This reduces the memory footprint and computational overhead while maintaining high accuracy. We validate our approach using the CIFAR-10 dataset. Our model achieves performance comparable to deep networks with significantly fewer parameters and lower computational complexity, demonstrating the potential of biologically inspired architectures for efficient edge AI.
title Early Exiting Predictive Coding Neural Networks for Edge AI
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2309.02022