Robustness analysis of Deep Sky Objects detection models on HPC

TL;DR

This paper evaluates the robustness of Deep Sky Objects detection models using HPC to parallelize computations, comparing models like YOLO and RET-DETR on telescope images for improved astronomical image analysis.

Contribution

It introduces a comprehensive robustness analysis of deep learning models for astronomical object detection using HPC, which is a novel approach in this domain.

Findings

HPC effectively accelerates robustness testing.

YOLO and RET-DETR show different robustness profiles.

Deep learning models can be optimized for astronomical imaging.

Abstract

Astronomical surveys and the growing involvement of amateur astronomers are producing more sky images than ever before, and this calls for automated processing methods that are accurate and robust. Detecting Deep Sky Objects -- such as galaxies, nebulae, and star clusters -- remains challenging because of their faint signals and complex backgrounds. Advances in Computer Vision and Deep Learning now make it possible to improve and automate this process. In this paper, we present the training and comparison of different detection models (YOLO, RET-DETR) on smart telescope images, using High-Performance Computing (HPC) to parallelise computations, in particular for robustness testing.