Contrastive learning of extragalactic stellar streams

Astronomy and Astrophysics, vol. 712, pp. A179

Abstract

We present a self-supervised approach for characterizing low-surface-brightness tidal features in wide-field imaging data by applying the nearest-neighbor contrastive learning of visual representations (NNCLR) algorithm to a curated subset of the Dark Energy Survey Data Release 2 (DES DR2). We constructed 38 334 cutouts of well-resolved galaxies in the g, r, i bands, applying a novel “tiered sigmoid scaling function” to dynamically adjust image contrast according to the object’s signal-to-noise (S/N) and background levels. A supplemental labeled sample of 366 galaxies enabled the qualitative assessment of the learned embeddings. We trained a convolutional neural network with image augmentations including injection of simulated background stars and projected the resulting 512-dimensional representations into two dimensions using uniform manifold approximation and projection (UMAP) and its local density preserving variant (densMAP). We find that the NNCLR latent space recovers global trends corresponding to major merger features, yet does not reliably separate stellar streams without further supervision. To interpret the network’s implicit attention, we computed gradient-based saliency maps averaged over the full dataset: these reveal that the tiered sigmoid scaling effectively attenuates information from the center of the image cutouts, thereby suppressing the learning of high-surface-brightness features of each image cutout’s central galaxy. Our study provides a blueprint for leveraging contrastive methods to mine forthcoming survey data for faint tidal substructure and highlights key preprocessing and interpretability considerations for robust stream detection.

Authors 4

1. Jelle Mes Aachen Leiden Observatory

   Leiden University · Leiden Observatory

   Affiliation as printed

   Leiden Observatory, Leiden University

2. Juan Miró-Carretero Aachen Leiden Observatory

   Universidad Complutense de Madrid · Leiden University · Leiden Observatory

   Affiliation as printed

   Departamento de Física de la Tierra y Astrofísica, Universidad Complutense de Madrid

   Leiden Observatory, Leiden University

3. Konrad H. Kuijken Aachen Leiden Observatory

   Leiden University · Leiden Observatory

   Affiliation as printed

   Leiden Observatory, Leiden University

4. Amina Helmi

   University of Groningen

   Affiliation as printed

   Kapteyn Astronomical Institute, University of Groningen