SpectraNet: Learned Recognition of Artificial Satellites From High Contrast Spectroscopic Imagery

TL;DR

This paper introduces a neural network-based method for identifying artificial satellites using spectroscopic data, achieving high accuracy even with re-orienting satellites, and incorporates uncertainty measurement techniques crucial for space traffic management.

Contribution

It presents a novel application of residual CNNs to distance-invariant spectroscopic satellite identification and demonstrates effective uncertainty quantification methods.

Findings

Over 80% classification accuracy in simulated 64-class problem.

72% accuracy achieved in real astronomical observations for 9-class problem.

Effective uncertainty estimation methods validated for space traffic decision-making.

Abstract

Effective space traffic management requires positive identification of artificial satellites. Current methods for extracting object identification from observed data require spatially resolved imagery which limits identification to objects in low earth orbits. Most artificial satellites, however, operate in geostationary orbits at distances which prohibit ground based observatories from resolving spatial information. This paper demonstrates an object identification solution leveraging modified residual convolutional neural networks to map distance-invariant spectroscopic data to object identity. We report classification accuracies exceeding 80% for a simulated 64-class satellite problem--even in the case of satellites undergoing constant, random re-orientation. An astronomical observing campaign driven by these results returned accuracies of 72% for a nine-class problem with an average of 100 examples per class, performing as expected from simulation. We demonstrate the application of variational Bayesian inference by dropout, stochastic weight averaging (SWA), and SWA-focused deep ensembling to measure classification uncertainties--critical components in space traffic management where routine decisions risk expensive space assets and carry geopolitical consequences.