Decentralised Semi-supervised Onboard Learning for Scene Classification in Low-Earth Orbit

Johan Östman, Pablo Gomez, Vinutha Magal Shreenath, Gabriele Meoni
5/6/2023
cs.LGcs.DCcs.MA

No official code found. Best implementation: gomezzz/DistMSMatch (0 stars, MIT, last commit 2025-03-05).

Abstract

Onboard machine learning on the latest satellite hardware offers the potential for significant savings in communication and operational costs. We showcase the training of a machine learning model on a satellite constellation for scene classification using semi-supervised learning while accounting for operational constraints such as temperature and limited power budgets based on satellite processor benchmarks of the neural network. We evaluate mission scenarios employing both decentralised and federated learning approaches. All scenarios achieve convergence to high accuracy (around 91% on EuroSAT RGB dataset) within a one-day mission timeframe.

View on arXivView PDF

Code Implementations(1)

gomezzz/DistMSMatchCommunity100%
MIT

A MPI-distributed version of MSMatch integrating the PASEOS framework for physical modelling

01PythonJun 21, 20221 years agoMIT

Similar Papers

Stitching Satellites to the Edge: Pervasive and Efficient Federated LEO Satellite Learning

85% match2024

Space for Improvement: Navigating the Design Space for Federated Learning in Satellite Constellations

83% match2024

FedSN: A Federated Learning Framework over Heterogeneous LEO Satellite Networks

83% match2023

Bringing Federated Learning to Space

83% match2025

Scheduling for On-Board Federated Learning with Satellite Clusters

81% match2024

Heterogeneity: An Open Challenge for Federated On-board Machine Learning

81% match2024

Cite this paper

@article{stman2023decentralised,
  title  = {Decentralised Semi-supervised Onboard Learning for Scene Classification in Low-Earth Orbit},
  author = {Johan Östman and Pablo Gomez and Vinutha Magal Shreenath and Gabriele Meoni},
  year   = {2023},
  eprint = {2305.04059},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2305.04059v1}
}

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/2305.04059.svg)](https://codeofpaper.com/paper/2305.04059)

Discussion