Fast automated analysis of strong gravitational lenses with convolutional neural networks
Abstract
Quantifying image distortions caused by strong gravitational lensing—the formation of multiple images of distant sources due to the deflection of their light by the gravity of intervening structures—and estimating the corresponding matter distribution of these structures (the ‘gravitational lens’) has primarily been performed using maximum likelihood modelling of observations. Our procedure is typically time- and resource-consuming, requiring sophisticated lensing codes, several data preparation steps, and finding the maximum likelihood model parameters in a computationally expensive process with downhill optimizers. Accurate analysis of a single gravitational lens can take up to a few weeks and requires expert knowledge of the physical processes and methods involved. Tens of thousands of new lenses are expected to be discovered with the upcoming generation of ground and space surveys. We report the use of deep convolutional neural networks to estimate lensing parameters in an extremely fast and automated way, circumventing the difficulties that are faced by maximum likelihood methods. We also show that the removal of lens light can be made fast and automated using independent component analysis of multi-filter imaging data. Our networks can recover the parameters of the ‘singular isothermal ellipsoid’ density profile, which is commonly used to model strong lensing systems, withmore »
- Authors:
-
- Stanford Univ., CA (United States). Kavli Inst. for Particle Astrophysics and Cosmology; SLAC National Accelerator Lab., Menlo Park, CA (United States)
- Publication Date:
- Research Org.:
- SLAC National Accelerator Laboratory (SLAC), Menlo Park, CA (United States)
- Sponsoring Org.:
- USDOE
- OSTI Identifier:
- 1390590
- Grant/Contract Number:
- AC02-76SF00515
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Nature (London)
- Additional Journal Information:
- Journal Name: Nature (London); Journal Volume: 548; Journal Issue: 7669; Journal ID: ISSN 0028-0836
- Publisher:
- Nature Publishing Group
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 79 ASTRONOMY AND ASTROPHYSICS; computational astrophysics; cosmology
Citation Formats
Hezaveh, Yashar D., Levasseur, Laurence Perreault, and Marshall, Philip J. Fast automated analysis of strong gravitational lenses with convolutional neural networks. United States: N. p., 2017.
Web. doi:10.1038/nature23463.
Hezaveh, Yashar D., Levasseur, Laurence Perreault, & Marshall, Philip J. Fast automated analysis of strong gravitational lenses with convolutional neural networks. United States. https://doi.org/10.1038/nature23463
Hezaveh, Yashar D., Levasseur, Laurence Perreault, and Marshall, Philip J. Wed .
"Fast automated analysis of strong gravitational lenses with convolutional neural networks". United States. https://doi.org/10.1038/nature23463. https://www.osti.gov/servlets/purl/1390590.
@article{osti_1390590,
title = {Fast automated analysis of strong gravitational lenses with convolutional neural networks},
author = {Hezaveh, Yashar D. and Levasseur, Laurence Perreault and Marshall, Philip J.},
abstractNote = {Quantifying image distortions caused by strong gravitational lensing—the formation of multiple images of distant sources due to the deflection of their light by the gravity of intervening structures—and estimating the corresponding matter distribution of these structures (the ‘gravitational lens’) has primarily been performed using maximum likelihood modelling of observations. Our procedure is typically time- and resource-consuming, requiring sophisticated lensing codes, several data preparation steps, and finding the maximum likelihood model parameters in a computationally expensive process with downhill optimizers. Accurate analysis of a single gravitational lens can take up to a few weeks and requires expert knowledge of the physical processes and methods involved. Tens of thousands of new lenses are expected to be discovered with the upcoming generation of ground and space surveys. We report the use of deep convolutional neural networks to estimate lensing parameters in an extremely fast and automated way, circumventing the difficulties that are faced by maximum likelihood methods. We also show that the removal of lens light can be made fast and automated using independent component analysis of multi-filter imaging data. Our networks can recover the parameters of the ‘singular isothermal ellipsoid’ density profile, which is commonly used to model strong lensing systems, with an accuracy comparable to the uncertainties of sophisticated models but about ten million times faster: 100 systems in approximately one second on a single graphics processing unit. These networks can provide a way for non-experts to obtain estimates of lensing parameters for large samples of data.},
doi = {10.1038/nature23463},
journal = {Nature (London)},
number = 7669,
volume = 548,
place = {United States},
year = {Wed Aug 30 00:00:00 EDT 2017},
month = {Wed Aug 30 00:00:00 EDT 2017}
}
Web of Science
Works referenced in this record:
The Third Gravitational Lensing Accuracy Testing (GREAT3) Challenge Handbook
text, January 2013
- Mandelbaum, Rachel; Rowe, Barnaby; Bosch, James
- arXiv
ImageNet Large Scale Visual Recognition Challenge
text, January 2015
- Jia, Deng,; Andrej, Karpathy,; Sean, Ma,
- The University of North Carolina at Chapel Hill University Libraries
Galaxy Zoo: morphological classifications for 120 000 galaxies in HST legacy imaging
journal, October 2016
- Willett, Kyle W.; Galloway, Melanie A.; Bamford, Steven P.
- Monthly Notices of the Royal Astronomical Society, Vol. 464, Issue 4
A Bayesian analysis of regularized source inversions in gravitational lensing
journal, September 2006
- Suyu, S. H.; Marshall, P. J.; Hobson, M. P.
- Monthly Notices of the Royal Astronomical Society, Vol. 371, Issue 2
Finding strong lenses in CFHTLS using convolutional neural networks
text, January 2017
- Jacobs, Colin; Glazebrook, Karl; Collett, Thomas
- arXiv
The sl2s Galaxy-Scale lens Sample. iii. lens Models, Surface Photometry, and Stellar Masses for the Final Sample
journal, October 2013
- Sonnenfeld, Alessandro; Gavazzi, Raphaël; Suyu, Sherry H.
- The Astrophysical Journal, Vol. 777, Issue 2
The population of galaxy-galaxy strong lenses in forthcoming optical imaging surveys
text, January 2015
- Collett, Thomas E.
- arXiv
ImageNet Large Scale Visual Recognition Challenge
journal, April 2015
- Russakovsky, Olga; Deng, Jia; Su, Hao
- International Journal of Computer Vision, Vol. 115, Issue 3
Approximation by superpositions of a sigmoidal function
journal, December 1989
- Cybenko, G.
- Mathematics of Control, Signals, and Systems, Vol. 2, Issue 4
Unreasonable effectiveness of learning neural networks: From accessible states and robust ensembles to basic algorithmic schemes
journal, November 2016
- Baldassi, Carlo; Borgs, Christian; Chayes, Jennifer T.
- Proceedings of the National Academy of Sciences, Vol. 113, Issue 48
The sl2s Galaxy-Scale lens Sample. v. dark Matter Halos and Stellar imf of Massive Early-Type Galaxies out to Redshift 0.8
journal, February 2015
- Sonnenfeld, Alessandro; Treu, Tommaso; Marshall, Philip J.
- The Astrophysical Journal, Vol. 800, Issue 2
A Bayesian analysis of regularized source inversions in gravitational lensing
journal, September 2006
- Suyu, S. H.; Marshall, P. J.; Hobson, M. P.
- Monthly Notices of the Royal Astronomical Society, Vol. 371, Issue 2
The SL2S Galaxy-scale Lens Sample. III. Lens Models, Surface Photometry and Stellar Masses for the final sample
text, January 2013
- Sonnenfeld, Alessandro; Gavazzi, Raphaël; Suyu, Sherry H.
- arXiv
Lens Models and Magnification maps of the six Hubble Frontier Fields Clusters
journal, November 2014
- Johnson, Traci L.; Sharon, Keren; Bayliss, Matthew B.
- The Astrophysical Journal, Vol. 797, Issue 1
The Third Gravitational Lensing Accuracy Testing (Great3) Challenge Handbook
journal, April 2014
- Mandelbaum, Rachel; Rowe, Barnaby; Bosch, James
- The Astrophysical Journal Supplement Series, Vol. 212, Issue 1
The SL2S Galaxy-scale Lens Sample. V. Dark Matter Halos and Stellar IMF of Massive Early-type Galaxies out to Redshift 0.8
text, January 2014
- Sonnenfeld, Alessandro; Treu, Tommaso; Marshall, Philip J.
- arXiv
Lens models and magnification maps of the six Hubble Frontier Fields clusters
text, January 2014
- Johnson, Traci L.; Sharon, Keren; Bayliss, Matthew B.
- arXiv
Finding strong lenses in CFHTLS using convolutional neural networks
journal, June 2017
- Jacobs, C.; Glazebrook, K.; Collett, T.
- Monthly Notices of the Royal Astronomical Society, Vol. 471, Issue 1
Galaxy Zoo: Morphological Classifications for 120,000 Galaxies in HST Legacy Imaging
text, January 2016
- Willett, Kyle W.; Galloway, Melanie A.; Bamford, Steven P.
- arXiv
The Sloan Lens ACS Survey. V. The Full ACS Strong‐Lens Sample
journal, August 2008
- Bolton, Adam S.; Burles, Scott; Koopmans, Léon V. E.
- The Astrophysical Journal, Vol. 682, Issue 2
A Bayesian analysis of regularised source inversions in gravitational lensing
text, January 2006
- Suyu, S. H.; Marshall, P. J.; Hobson, M. P.
- arXiv
Unreasonable Effectiveness of Learning Neural Networks: From Accessible States and Robust Ensembles to Basic Algorithmic Schemes
text, January 2016
- Baldassi, Carlo; Borgs, Christian; Chayes, Jennifer
- arXiv
A Systematic Review of Strong Gravitational Lens Modeling Software
text, January 2012
- Lefor, Alan T.; Futamase, Toshifumi; Akhlaghi, Mohammad
- arXiv
Semilinear Gravitational Lens Inversion
journal, June 2003
- Warren, S. J.; Dye, S.
- The Astrophysical Journal, Vol. 590, Issue 2
Approximation by superpositions of a sigmoidal function
journal, December 1992
- Cybenko, G.
- Mathematics of Control, Signals, and Systems, Vol. 5, Issue 4
A systematic review of strong gravitational lens modeling software
journal, July 2013
- Lefor, Alan T.; Futamase, Toshifumi; Akhlaghi, Mohammad
- New Astronomy Reviews, Vol. 57, Issue 1-2
The Population of Galaxy–Galaxy Strong Lenses in Forthcoming Optical Imaging Surveys
journal, September 2015
- Collett, Thomas E.
- The Astrophysical Journal, Vol. 811, Issue 1
CMU DeepLens: deep learning for automatic image-based galaxy–galaxy strong lens finding
journal, July 2017
- Lanusse, François; Ma, Quanbin; Li, Nan
- Monthly Notices of the Royal Astronomical Society, Vol. 473, Issue 3
Enabling Dark Energy Science with Deep Generative Models of Galaxy Images
journal, February 2017
- Ravanbakhsh, Siamak; Lanusse, Francois; Mandelbaum, Rachel
- Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 31, Issue 1
Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning
journal, February 2017
- Szegedy, Christian; Ioffe, Sergey; Vanhoucke, Vincent
- Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 31, Issue 1
Works referencing / citing this record:
Is every strong lens model unhappy in its own way? Uniform modelling of a sample of 13 quadruply+ imaged quasars
text, January 2018
- Shajib, A. J.; Birrer, S.; Treu, T.
- arXiv
Probing the universe with citizen science and gravitational lensing
dissertation, January 2017
- Rafael, Küng,
- University of Zurich
Deep learning predictions of galaxy merger stage and the importance of observational realism
journal, October 2019
- Bottrell, Connor; Hani, Maan H.; Teimoorinia, Hossen
- Monthly Notices of the Royal Astronomical Society, Vol. 490, Issue 4
Response to NITRD, NCO, NSF Request for Information on "Update to the 2016 National Artificial Intelligence Research and Development Strategic Plan"
preprint, January 2019
- Amundson, J.; Annis, J.; Avestruz, C.
- arXiv
Neural network-based anomaly detection for high-resolution X-ray spectroscopy
journal, June 2019
- Ichinohe, Y.; Yamada, S.
- Monthly Notices of the Royal Astronomical Society, Vol. 487, Issue 2
Deep-CEE I: fishing for galaxy clusters with deep neural nets
journal, October 2019
- Chan, Matthew C.; Stott, John P.
- Monthly Notices of the Royal Astronomical Society, Vol. 490, Issue 4
X-ray Study of Spatial Structures in Tycho's Supernova Remnant Using Unsupervised Deep Learning
text, January 2019
- Iwasaki, Hiroyoshi; Ichinohe, Yuto; Uchiyama, Yasunobu
- arXiv
Deep learning predictions of galaxy merger stage and the importance of observational realism
text, January 2019
- Bottrell, Connor; Hani, Maan H.; Teimoorinia, Hossen
- arXiv
Searching for Subsecond Stellar Variability with Wide-field Star Trails and Deep Learning
journal, November 2018
- Thomas, David; Kahn, Steven M.
- The Astrophysical Journal, Vol. 868, Issue 1
Galaxy Zoo: probabilistic morphology through Bayesian CNNs and active learning
journal, October 2019
- Walmsley, Mike; Smith, Lewis; Lintott, Chris
- Monthly Notices of the Royal Astronomical Society, Vol. 491, Issue 2
Optical Deep Learning Nano-Profilometry
text, January 2019
- Zhu, Jinlong; Liu, Yanan; Purandare, Sanyogita
- Unpublished
Finding high-redshift strong lenses in DES using convolutional neural networks
text, January 2018
- Jacobs, C.; Collett, T.; Glazebrook, K.
- arXiv
Wat zei je? Detecting Out-of-Distribution Translations with Variational Transformers
preprint, January 2020
- Xiao, Tim Z.; Gomez, Aidan N.; Gal, Yarin
- arXiv
Learning to Predict the Cosmological Structure Formation
text, January 2018
- He, Siyu; Li, Yin; Feng, Yu
- arXiv
Neural network-based preprocessing to estimate the parameters of the X-ray emission of a single-temperature thermal plasma
text, January 2018
- Ichinohe, Y.; Yamada, S.; Miyazaki, N.
- arXiv
Using convolutional neural networks to predict galaxy metallicity from three-colour images
journal, February 2019
- Wu, John F.; Boada, Steven
- Monthly Notices of the Royal Astronomical Society, Vol. 484, Issue 4
Reconstructing Functions and Estimating Parameters with Artificial Neural Networks: A Test with the Hubble Parameter and SNe Ia
text, January 2019
- Wang, Guo-Jian; Ma, Xiao-Jiao; Li, Si-Yao
- arXiv
Radio Galaxy Zoo: Claran – a deep learning classifier for radio morphologies
journal, October 2018
- Wu, Chen; Wong, Oiwei Ivy; Rudnick, Lawrence
- Monthly Notices of the Royal Astronomical Society, Vol. 482, Issue 1
Discovery of CDK4 inhibitors by convolutional neural networks
journal, February 2019
- Xu, Yinqiu; Chen, Pingping; Lin, Xinhao
- Future Medicinal Chemistry, Vol. 11, Issue 3
A Transdisciplinary Review of Deep Learning Research and Its Relevance for Water Resources Scientists
journal, November 2018
- Shen, Chaopeng
- Water Resources Research, Vol. 54, Issue 11
Phased microphone array for sound source localization with deep learning
journal, May 2019
- Ma, Wei; Liu, Xun
- Aerospace Systems, Vol. 2, Issue 2
Lessons from a blind study of simulated lenses: image reconstructions do not always reproduce true convergence
text, January 2019
- Denzel, Philipp; Mukherjee, Sampath; Coles, Jonathan P.
- arXiv
A trans-disciplinary review of deep learning research for water resources scientists
text, January 2017
- Shen, Chaopeng
- arXiv
Automated Assignment of Rotational Spectra Using Artificial Neural Networks
conference, June 2018
- Zaleski, Daniel; Prozument, Kirill
- Proceedings of the 73rd International Symposium on Molecular Spectroscopy
Mining for Dark Matter Substructure: Inferring Subhalo Population Properties from Strong Lenses with Machine Learning
journal, November 2019
- Brehmer, Johann; Mishra-Sharma, Siddharth; Hermans, Joeri
- The Astrophysical Journal, Vol. 886, Issue 1
Deep learning reconstruction of ultrashort pulses
journal, January 2018
- Zahavy, Tom; Dikopoltsev, Alex; Moss, Daniel
- Optica, Vol. 5, Issue 5
Image Simulations for Strong and Weak Gravitational Lensing
journal, March 2020
- A. Plazas, Andrés
- Symmetry, Vol. 12, Issue 4
A Bayesian neural network predicts the dissolution of compact planetary systems
text, January 2021
- Cranmer, Miles; Tamayo, Daniel; Rein, Hanno
- arXiv
Neural network-based preprocessing to estimate the parameters of the X-ray emission of a single-temperature thermal plasma
journal, February 2018
- Ichinohe, Y.; Yamada, S.; Miyazaki, N.
- Monthly Notices of the Royal Astronomical Society, Vol. 475, Issue 4
Galaxy Zoo: Probabilistic Morphology through Bayesian CNNs and Active Learning
text, January 2019
- Walmsley, Mike; Smith, Lewis; Lintott, Chris
- arXiv
Learning to predict the cosmological structure formation
journal, June 2019
- He, Siyu; Li, Yin; Feng, Yu
- Proceedings of the National Academy of Sciences, Vol. 116, Issue 28
Machine Learning Detection of Bell Nonlocality in Quantum Many-Body Systems
journal, June 2018
- Deng, Dong-Ling
- Physical Review Letters, Vol. 120, Issue 24
A Modular Deep Learning Pipeline for Galaxy-Scale Strong Gravitational Lens Detection and Modeling
preprint, January 2019
- Madireddy, Sandeep; Ramachandra, Nesar; Li, Nan
- arXiv
Seeing permeability from images: fast prediction with convolutional neural networks
journal, September 2018
- Wu, Jinlong; Yin, Xiaolong; Xiao, Heng
- Science Bulletin, Vol. 63, Issue 18
Phased microphone array for sound source localization with deep learning
journal, May 2019
- Ma, Wei; Liu, Xun
- Aerospace Systems, Vol. 2, Issue 2
Lenstronomy: multi-purpose gravitational lens modelling software package
text, January 2018
- Birrer, Simon; Amara, Adam
- arXiv
Machine learning and the physical sciences
journal, December 2019
- Carleo, Giuseppe; Cirac, Ignacio; Cranmer, Kyle
- Reviews of Modern Physics, Vol. 91, Issue 4
Deep Learning Reconstruction of Ultra-Short Pulses
preprint, January 2018
- Zahavy, Tom; Dikopoltsev, Alex; Cohen, Oren
- arXiv
A new era in the search for dark matter
journal, October 2018
- Bertone, Gianfranco; Tait, Tim M. P.
- Nature, Vol. 562, Issue 7725
Lessons from a blind study of simulated lenses: image reconstructions do not always reproduce true convergence
journal, January 2020
- Denzel, Philipp; Mukherjee, Sampath; Coles, Jonathan P.
- Monthly Notices of the Royal Astronomical Society, Vol. 492, Issue 3
Direct detection of dark matter substructure in strong lens images with convolutional neural networks
journal, January 2020
- Diaz Rivero, Ana; Dvorkin, Cora
- Physical Review D, Vol. 101, Issue 2
SWIRL: High-performance many-core CPU code generation for deep neural networks
journal, May 2019
- Venkat, Anand; Rusira, Tharindu; Barik, Raj
- The International Journal of High Performance Computing Applications, Vol. 33, Issue 6
An Extended Catalog of Galaxy–Galaxy Strong Gravitational Lenses Discovered in DES Using Convolutional Neural Networks
journal, July 2019
- Jacobs, C.; Collett, T.; Glazebrook, K.
- The Astrophysical Journal Supplement Series, Vol. 243, Issue 1
Learning to predict the cosmological structure formation
journal, June 2019
- He, Siyu; Li, Yin; Feng, Yu
- Proceedings of the National Academy of Sciences, Vol. 116, Issue 28
Machine learning meets quantum foundations: A brief survey
journal, October 2020
- Bharti, Kishor; Haug, Tobias; Vedral, Vlatko
- AVS Quantum Science, Vol. 2, Issue 3
Machine Learning Bell Nonlocality in Quantum Many-body Systems
text, January 2017
- Deng, Dong-Ling
- arXiv
Machine Learning for Nanohertz Gravitational Wave Detection and Parameter Estimation with Pulsar Timing Array
preprint, January 2020
- Chen, MengNi; Zhong, Yuanhong; Feng, Yi
- arXiv
Regularized pseudo-phase imaging for inspecting and sensing nanoscale features
journal, January 2019
- Zhu, Jinlong; Zhou, Renjie; Zhang, Lenan
- Optics Express, Vol. 27, Issue 5
Automated assignment of rotational spectra using artificial neural networks
journal, September 2018
- Zaleski, Daniel P.; Prozument, Kirill
- The Journal of Chemical Physics, Vol. 149, Issue 10
Genetic-algorithm-optimized neural networks for gravitational wave classification
text, January 2020
- Deighan, Dwyer S.; Field, Scott E.; Capano, Collin D.
- arXiv
X-ray study of spatial structures in Tycho’s supernova remnant using unsupervised deep learning
journal, July 2019
- Iwasaki, Hiroyoshi; Ichinohe, Yuto; Uchiyama, Yasunobu
- Monthly Notices of the Royal Astronomical Society, Vol. 488, Issue 3
Dark Matter Subhalos, Strong Lensing and Machine Learning
preprint, January 2020
- Varma, Sreedevi; Fairbairn, Malcolm; Figueroa, Julio
- arXiv
Nuclear liquid-gas phase transition with machine learning
journal, November 2020
- Wang, Rui; Ma, Yu-Gang; Wada, R.
- Physical Review Research, Vol. 2, Issue 4
Phased Microphone Array for Sound Source Localization with Deep Learning
preprint, January 2018
- Ma, Wei; Liu, Xun
- arXiv
Scientific intuition inspired by machine learning-generated hypotheses
journal, April 2021
- Friederich, Pascal; Krenn, Mario; Tamblyn, Isaac
- Machine Learning: Science and Technology, Vol. 2, Issue 2
Is every strong lens model unhappy in its own way? Uniform modelling of a sample of 13 quadruply+ imaged quasars
journal, December 2018
- Shajib, A. J.; Birrer, S.; Treu, T.
- Monthly Notices of the Royal Astronomical Society, Vol. 483, Issue 4
Finding high-redshift strong lenses in DES using convolutional neural networks
journal, January 2019
- Jacobs, C.; Collett, T.; Glazebrook, K.
- Monthly Notices of the Royal Astronomical Society, Vol. 484, Issue 4
Data-driven Reconstruction of Gravitationally Lensed Galaxies Using Recurrent Inference Machines
journal, September 2019
- Morningstar, Warren R.; Levasseur, Laurence Perreault; Hezaveh, Yashar D.
- The Astrophysical Journal, Vol. 883, Issue 1
Reconstructing Functions and Estimating Parameters with Artificial Neural Networks: A Test with a Hubble Parameter and SNe Ia
journal, January 2020
- Wang, Guo-Jian; Ma, Xiao-Jiao; Li, Si-Yao
- The Astrophysical Journal Supplement Series, Vol. 246, Issue 1
Direct Detection of Dark Matter Substructure in Strong Lens Images with Convolutional Neural Networks
text, January 2019
- Rivero, Ana Diaz; Dvorkin, Cora
- arXiv
Hunting for Dark Matter Subhalos in Strong Gravitational Lensing with Neural Networks
preprint, January 2020
- Lin, Joshua Yao-Yu; Yu, Hang; Morningstar, Warren
- arXiv
Quasar microlensing light-curve analysis using deep machine learning
journal, April 2019
- Vernardos, Georgios; Tsagkatakis, Grigorios
- Monthly Notices of the Royal Astronomical Society, Vol. 486, Issue 2
Fast point spread function modeling with deep learning
journal, July 2018
- Herbel, Jörg; Kacprzak, Tomasz; Amara, Adam
- Journal of Cosmology and Astroparticle Physics, Vol. 2018, Issue 07