Julia Dima | Searcharxiv

arXiv2026

High-Redshift Gravitational Lens Discoveries in JWST NIRCam Using AnomalyMatch

Context. Strong gravitational lenses provide a unique tool to probe cosmology and astrophysics at high redshift, offering constraints on the mass distribution of background source populations. Despite their scientific value, their rarity and subtle visual features make them challenging to identify in the wealth of data delivered by facilities such as the James Webb Space Telescope (JWST), whose unmatched resolution and near-infrared coverage make it particularly well-suited to detecting lensing systems in this regime. Aims. We make use of the specialised open-source software AnomalyMatch, a semi-supervised learning method to trawl the ASTRODEEP and COSMOS-Web surveys for gravitational lenses. Methods. Building on a training dataset of eleven previously identified gravitational lenses, we use AnomalyMatch and its iterative human-in-the-loop method to train a neural network to identify gravitational lenses in JWST Level 3 products using ESA Datalabs. Results. In total we identify 58 unique gravitational lenses. These are graded by four experts into 16 Grade A, 16 Grade B, and 26 Grade C lenses. Of all lenses identified, 37 were previously uncatalogued. We analyse their properties such as photometric redshift measurements and spectroscopic redshift, when the latter is available. The lenses previously identified span spectroscopic redshifts to zspec < 1.39 and photometric redshifts to zphot < 2.21. The uncatalogued lens system with the highest redshift is at zphot = 2.1. Conclusions. Overall, we demonstrate the potential of AnomalyMatch for large-scale searches for gravitational lenses and other rare high-redshift objects in JWST archives.

astro-ph.GA↗

arXiv2024

XAMI -- A Benchmark Dataset for Artefact Detection in XMM-Newton Optical Images

Reflected or scattered light produce artefacts in astronomical observations that can negatively impact the scientific study. Hence, automated detection of these artefacts is highly beneficial, especially with the increasing amounts of data gathered. Machine learning methods are well-suited to this problem, but currently there is a lack of annotated data to train such approaches to detect artefacts in astronomical observations. In this work, we present a dataset of images from the XMM-Newton space telescope Optical Monitoring camera showing different types of artefacts. We hand-annotated a sample of 1000 images with artefacts which we use to train automated ML methods. We further demonstrate techniques tailored for accurate detection and masking of artefacts using instance segmentation. We adopt a hybrid approach, combining knowledge from both convolutional neural networks (CNNs) and transformer-based models and use their advantages in segmentation. The presented method and dataset will advance artefact detection in astronomical observations by providing a reproducible baseline. All code and data are made available (https://github.com/ESA-Datalabs/XAMI-model and https://github.com/ESA-Datalabs/XAMI-dataset).

cs.CV↗