Star localization in astronomical imagery is a computer vision task that underpins satellite tracking. Astronomical star extraction techniques often struggle to detect stars when applied to satellite tracking imagery due to the narrower fields of view and rate track observational modes of satellite tracking telescopes. We present a large dataset of real narrow-field rate-tracked imagery with ground truth stars, created using a combination of existing star detection techniques, an astrometric engine, and a star catalog. We train three state of the art object detection, instance segmentation, and line segment detection models on this dataset and evaluate them with object-wise, pixel-wise, and astrometric metrics. Our proposed approaches require no metadata; when paired with a lost-in-space astrometric engine, they find astrometric fits based solely on uncorrected image pixels. Experimental results on real data indicate the effectiveness of learned star detection: we report astrometric fit rates over double that of classical star detection algorithms, improved dim star recall, and comparable star localization residuals.

An emerging class of Fizeau optical telescopes have the potential to upend prior cost scaling models, substantially improving the angular resolution and contrast attainable by ground-based astronomical instruments. However, this design introduces a challenging visual control problem that must be solved to compensate for wavefront aberrations induced by the flexible substructure it employs. We subvert this problem with a deep optics approach to policy design and image recovery that exploits, rather than corrects, aberrations to obtain domain-specific object recovery performance exceeding that of more costly filled aperture designs.

In this work, we propose to consolidate radio frequency communication signals and speech audio into a common data modality: multichannel, time-continuous amplitudes with characteristic spectrograms and a finite symbol alphabet. By putting a portion of the radio spectrum on a similar footing to audio, this may allow us to leverage a great deal of the technological progress achieved by automatic speech recognition (ASR) and readily transfer it to radio frequency machine learning (RFML), a rapidly developing field. To support this claim, we take the leading ASR architecture of wav2vec2 and apply it directly to a challenging dataset of real, low-SNR radio signals captured from satellite telecommunications. Representing the first large-scale application of learned detection and classification of raw signals emitted from a diverse array of active low Earth orbit satellites, the speech-inspired network demonstrates strong proficiency on all tasks and robustness to the degraded signal environment.

In the space domain, remotely measuring the rotational state of a spacecraft provides critical information for assessing its operational health. For the large family of space assets that lie in regions of space too distant to resolve spatial features with ground sensors, measurement of the energy spectrum of reflected sunlight has shown recent promise in probing their material and spatial properties. In this work, we explore the utility of using the recently proposed Implicit-PDF network to implicitly learn challenging probability distributions associated with satellite orientations using only raw optical spectra as input. Originally designed for computer vision applications, we detail how the Implicit-PDF architecture can be applied directly to spectra and further extended to a broad class of signal inputs. We discuss key aspects to improving performance and demonstrate the generalizability of the implicit framework by showing how a single network can efficiently learn mutually exclusive probability distributions over multiple satellite classes without meaningful performance loss.1

Effective space domain awareness 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. Many artificial Earth 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 domain awareness where routine decisions risk expensive space assets and carry geopolitical consequences.

The detection of dim artificial Earth satellites using ground-based electro-optical sensors, particularly in the presence of background light, is technologically challenging. This perceptual task is foundational to our understanding of the space environment, and grows in importance as the number, variety, and dynamism of space objects increases. We present a hybrid image- and event-based architecture that leverages dynamic vision sensing technology to detect resident space objects in geosynchronous Earth orbit. Given the asynchronous, one-dimensional image data supplied by a dynamic vision sensor, our architecture applies conventional image feature extractors to integrated, two-dimensional frames in conjunction with point-cloud feature extractors, such as PointNet, in order to increase detection performance for dim objects in scenes with high background activity. In addition, an end-to-end event-based imaging simulator is developed to both produce data for model training as well as approximate the optimal sensor parameters for event-based sensing in the context of electrooptical telescope imagery. Experimental results confirm that the inclusion of point-cloud feature extractors increases recall for dim objects in the high-background regime.

A constructive learning algorithm dynamically creates a problem-specific neural network architecture rather than learning on a pre-specified architecture. The authors propose a parallel version of their recently presented constructive neural network learning algorithm. Parallelization provides a computational speedup by a factor of O(t) where t is the number of training examples. Distributed and parallel implementations under p4 using a network of workstations and a Touchstone DELTA are examined. Experimental results indicate that algorithm parallelization may result not only in improved computational time, but also in better prediction quality.>