Probabilistic segmentation of overlapping galaxies for large cosmological surveys

TL;DR

This paper explores a probabilistic U-Net model for galaxy segmentation in large surveys, emphasizing uncertainty quantification in overlapping galaxy images, which is crucial for precise cosmological analysis.

Contribution

It adapts the probabilistic U-Net for galaxy deblending, demonstrating effective uncertainty estimation from limited ground truth data in astrophysics.

Findings

Model captures pixel-wise uncertainty in galaxy segmentation.

Uncertainty propagation benefits galaxy property analysis.

First application of probabilistic segmentation for galaxy deblending.

Abstract

Encoder-Decoder networks such as U-Nets have been applied successfully in a wide range of computer vision tasks, especially for image segmentation of different flavours across different fields. Nevertheless, most applications lack of a satisfying quantification of the uncertainty of the prediction. Yet, a well calibrated segmentation uncertainty can be a key element for scientific applications such as precision cosmology. In this on-going work, we explore the use of the probabilistic version of the U-Net, recently proposed by Kohl et al (2018), and adapt it to automate the segmentation of galaxies for large photometric surveys. We focus especially on the probabilistic segmentation of overlapping galaxies, also known as blending. We show that, even when training with a single ground truth per input sample, the model manages to properly capture a pixel-wise uncertainty on the segmentation map. Such uncertainty can then be propagated further down the analysis of the galaxy properties. To our knowledge, this is the first time such an experiment is applied for galaxy deblending in astrophysics.