Hostname: page-component-586b7cd67f-dlnhk Total loading time: 0 Render date: 2024-11-28T03:06:00.639Z Has data issue: false hasContentIssue false
  • English
  • Français

Strong Lens Detection 2.0: Machine Learning and Transformer Models

Published online by Cambridge University Press:  04 March 2024

Hareesh Thuruthipilly Open the ORCID record for Hareesh Thuruthipilly [Opens in a new window]
Hareesh Thuruthipilly*
Affiliation:
National Centre for Nuclear Research, Astrophysics division ul. Pasteura 7, 02-093 Warszawa
*
email: [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Core share and HTML view are not available for this content. However, as you have access to this content, a full PDF is available via the ‘Save PDF’ action button.

Upcoming large-scale surveys like LSST are expected to uncover approximately 105 strong gravitational lenses within massive datasets. Traditional manual techniques are too time-consuming and impractical for such volumes of data. Consequently, machine learning methods have emerged as an alternative. In our prior work (Thuruthipilly et al. 2022), we introduced a self-attention-based machine learning model (transformers) for detecting strong gravitational lenses in simulated data from the Bologna Lens Challenge. These models offer advantages over simpler convolutional neural networks (CNNs) and competitive performance compared to state-of-the-art CNN models. We applied this model to the datasets from Bologna Lens Challenge 1 and 2 and simulated data on Euclid.

Keywords

Gravitational strong lensingdata analysisimage processingobservations
Type
Poster Paper
Information
Proceedings of the International Astronomical Union , Volume 18 , Symposium S381: Strong Gravitational Lensing in the Era of Big Data , December 2022 , pp. 28 - 30
Check for updates
Copyright
© The Author(s), 2024. Published by Cambridge University Press on behalf of International Astronomical Union

References

Bom, et al. 2022, Developing a victorious strategy to the second strong gravitational lensing data challenge. MNRAS, 515(4), 5121–5134.CrossRefGoogle Scholar
Euclid Collaboration 2023, Euclid Preparation TBD. Characterization of convolutional neural networks for the identification of galaxy-galaxy strong lensing events. arXiv e-prints, arXiv:2307.08736.Google Scholar
Hewitt, J. N., Turner, E. L., Schneider, D. P., Burke, B. F., & Langston, G. I. 1988, Unusual radio source MG1131+0456: a possible Einstein ring. Natur, 333(6173), 537–540.CrossRefGoogle Scholar
Ivezić, Ž., Kahn, S. M., Tyson, J. A., Abel, B., Acosta, E., Allsman, R., Alonso, D., AlSayyad, Y., Anderson, S. F., Andrew, J., & et al. 2019, Lsst: From science drivers to reference design and anticipated data products. ApJ, 873(2), 111.CrossRefGoogle Scholar
Metcalf, R. B., Meneghetti, M., Avestruz, C., Bellagamba, F., Bom, C. R., Bertin, E., Cabanac, R., Courbin, F., Davies, A., Decencière, E., & et al. 2019, The strong gravitational lens finding challenge. Astronomy & Astrophysics, 625, A119.CrossRefGoogle Scholar
R., S., J., A., Y., M., C., B., & S., C. C. 2021, Euclid preparation: I. The Euclid Wide Survey. arXiv e-prints, arXiv:2108.01201.Google Scholar
Thuruthipilly, H., Zadrozny, A., Pollo, A., & Biesiada, M. 2022, Finding strong gravitational lenses through self-attention. Study based on the Bologna Lens Challenge. Astronomy & Astrophysics, 664, A4.CrossRefGoogle Scholar
Vaswani, A. Attention is all you need. In Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, December 4-9, 2017, Long Beach, CA, USA 2017, pp. 5998–6008.Google Scholar