Detection of Strongly Lensed Arcs in Galaxy Clusters with Transformers

Abstract

Strong lensing in galaxy clusters probes properties of dense cores of dark matter halos in mass, studies the distant universe at flux levels and spatial resolutions otherwise unavailable, and constrains cosmological models independently. The next-generation large-scale sky imaging surveys are expected to discover thousands of cluster-scale strong lenses, which would lead to unprecedented opportunities for applying cluster-scale strong lenses to solve astrophysical and cosmological problems. However, the large data set challenges astronomers to identify and extract strong-lensing signals, particularly strongly lensed arcs, because of their complexity and variety. Hence, we propose a framework to detect cluster-scale strongly lensed arcs, which contains a transformer-based detection algorithm and an image simulation algorithm. We embed prior information of strongly lensed arcs at cluster scale into the training data through simulation and then train the detection algorithm with simulated images. We use the trained transformer to detect strongly lensed arcs from simulated and real data. Results show that our approach could achieve 99.63% accuracy rate, 90.32% recall rate, 85.37% precision rate, and 0.23% false-positive rate in detection of strongly lensed arcs from simulated images and could detect almost all strongly lensed arcs in real observation images. Besides, with an interpretation method, we have shown that our method could identify important information embedded in simulated data. Next, to test the reliability and usability of our approach, we will apply it to available observations (e.g., DESI Legacy Imaging Surveys 6 6 https://www.legacysurvey.org/ ) and simulated data of upcoming large-scale sky surveys, such as Euclid 7 7 https://www.euclid-ec.org/ and the China Space Station Telescope. 8 8 https://nao.cas.cn/csst/

CSST Strong-lensing Preparation: A Framework for Detecting Strong Lenses in the Multicolor Imaging Survey by the China Survey Space Telescope (CSST)

“…Generating simulated data, as discussed in our previous papers (Jia et al 2022(Jia et al , 2023a, serves a crucial purpose in the detection of celestial objects, particularly when sufficient training data are lacking. Simulated data provide valuable prior information about the targets we aim to detect.…”

“…These methods provided approaches for strong-lensing system finding and classification in Euclid data. Jia et al (2022) have introduced a transformer-based deep neural network (DETR) for strong-lensing system detection, which demonstrated noteworthy proficiency in identifying strong-lensing systems on the scale of galaxy clusters. The transformer model is designed to concentrate its attention on key regions like distorted images or arcs during the process of strong-lensing system detection.…”

PNet—A Deep Learning Based Photometry and Astrometry Bayesian Framework

“…In this study, our focus is on detecting point-like celestial objects in sparse star fields and extracting their positions and magnitudes, as this is a crucial prerequisite for studying these objects in time-domain astronomy. It is worth noting that extended targets and dense star fields (the distance between stars are less than twice the full width at half magnitude of the point-spread function, PSF) may benefit from multicolor images and other relevant neural networks for better detection and classification (González et al 2018;Farias et al 2020;Cheng et al 2021;Jia et al 2022;Yu et al 2022;Jia et al 2023b;Andrew 2023), or from methods specifically designed for dense star fields (Liu et al 2021a;Hansen et al 2022). Moreover, transients, such as supernovae in galaxies, can be further processed using methods based on the image difference or techniques developed based on temporal sequences of images (Kessler et al 2015;Wright et al 2015;Zackay et al 2016;Sánchez et al 2019;Gómez et al 2020;Mong et al 2020;Hu et al 2022;Makhlouf et al 2022).…”