Fast model inference and training on-board of Satellites

Vít Růžička, Gonzalo Mateo-García, Chris Bridges, Chris Brunskill, Cormac Purcell +2 more
7/17/2023
cs.AIcs.CV

No official code found. Best implementation: previtus/RaVAEn-unibap-dorbit (3 stars, last commit 2023-10-02).

Abstract

Artificial intelligence onboard satellites has the potential to reduce data transmission requirements, enable real-time decision-making and collaboration within constellations. This study deploys a lightweight foundational model called RaVAEn on D-Orbit's ION SCV004 satellite. RaVAEn is a variational auto-encoder (VAE) that generates compressed latent vectors from small image tiles, enabling several downstream tasks. In this work we demonstrate the reliable use of RaVAEn onboard a satellite, achieving an encoding time of 0.110s for tiles of a 4.8x4.8 km$^2$ area. In addition, we showcase fast few-shot training onboard a satellite using the latent representation of data. We compare the deployment of the model on the on-board CPU and on the available Myriad vision processing unit (VPU) accelerator. To our knowledge, this work shows for the first time the deployment of a multi-task model on-board a CubeSat and the on-board training of a machine learning model.

View on arXivView PDF

Code Implementations(1)

previtus/RaVAEn-unibap-dorbitCommunity100%

Running the RaVAEn model on board of D-Orbit satellites

31Shell, PythonNov 26, 20223 years ago

Similar Papers

Lightweight, Pre-trained Transformers for Remote Sensing Timeseries

83% match1 repo2792023

On-orbit model training for satellite imagery with label proportions

83% match1 repo32023

Onboard Satellite Image Classification for Earth Observation: A Comparative Study of ViT Models

82% match1 repo102024

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

82% match2024

Characterizing Satellite Geometry via Accelerated 3D Gaussian Splatting

82% match2024

TinyML Enhances CubeSat Mission Capabilities

82% match2026

Cite this paper

@article{rika2023fast,
  title  = {Fast model inference and training on-board of Satellites},
  author = {Vít Růžička and Gonzalo Mateo-García and Chris Bridges and Chris Brunskill and Cormac Purcell and Nicolas Longépé and Andrew Markham},
  year   = {2023},
  eprint = {2307.08700},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2307.08700v1}
}

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/2307.08700.svg)](https://codeofpaper.com/paper/2307.08700)

Discussion