Euclid Quick Data Release (Q1). Active galactic nuclei identification using diffusion-based inpainting of Euclid VIS images

Light emission from galaxies exhibit diverse brightness profiles, influenced by factors such as galaxy type, structural features, and interactions with other galaxies. Elliptical galaxies feature more uniform light distributions, while spiral and irregular galaxies have complex, varied light profiles due to their structural heterogeneity and star-forming activity. In addition, galaxies with active galactic nuclei (AGN) feature intense, concentrated emission from gas accretion around supermassive black holes, superimposed on regular galactic light, while quasi-stellar objects (QSOs) represent extreme cases in which AGN emissions dominate their host galaxies. The challenge of identifying AGN and QSOs has been discussed many times in the literature, often requiring multi-wavelength observations. This paper introduces a novel approach to identify AGN and QSOs from a single image. Diffusion models have recently been developed in the machine-learning literature to generate realistic-looking images of everyday objects. Utilising the spatial resolving power of the Euclid VIS images, we created a diffusion model trained on one million sources, without using any source pre-selection or labels. The model learns to reconstruct light distributions of normal galaxies, since the population is dominated by them. We conditioned the prediction of the central light distribution by masking the central few pixels of each source and reconstructed the light according to the diffusion model. We further used this prediction to identify sources that deviate from this profile by examining the reconstruction error of the few central pixels regenerated in each source's core. Our approach, solely using VIS imaging, features high completeness compared to traditional methods of AGN and QSO selection, including optical, near-infrared, mid-infrared, and X-rays. Our study offers practical insights for refining diffusion models and broadening their applications throughout the Euclid survey area, underscoring the utility of this approach in diverse astronomical contexts beyond just AGN identification.

Authors

Publication Details

Journal
Astronomy and Astrophysics
Published
2025-10-17
DOI
https://doi.org/10.1051/0004-6361/202554612
Primary Topic
Medical Image Segmentation Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Euclid Quick Data Release (Q1). Active galactic nuclei identification using diffusion-based inpainting of Euclid VIS images

K. George, Laura Bisigello, G. Rodighiero, P.-A. Duc et al.
Astronomy and Astrophysics
Medical Image Segmentation Techniques
article

Euclid Quick Data Release (Q1). Active galactic nuclei identification using diffusion-based inpainting of Euclid VIS images

K. George, Laura Bisigello, G. Rodighiero, P.-A. Duc, F Passalacqua, Will J. Percival, M. Ballardini, G. Mainetti, R. P. Saglia, P. Tallada-Crespí, S Sacquegna, G. Riccio, A. Pisani, D. Le Mignant, C. Sirignano, M. Brescia, E. Gaztanaga, E. Merlin, I. Hook, F. Dubath, C. Baccigalupi, M. Bolzonella, J. Carretero, M. Schirmer, J. Steinwagner, D. B. Sanders, F. Marulli, J. García-Bellido, S. Borgani, A.G Ferrari, A. Navarro-Alsina, C. Giocoli, M. Mezcua, A. Franco, S. A. Stanford, Grant M. Stevens, Patrice Simon, F. Courbin, V. Scottez, C. Valieri, Alex D. Hall, A. Kiessling, A. Troja, Michael D. Brown, F. Grupp, A. Grazian, M. Jhabvala, F. Torradeflot, Elena Sellentin, A. Díaz‐Sánchez, J.J Diaz, Pierluigi Monaco, S. Bardelli, G. Desprez, Daniele Bertacca, M. Radovich, M. Martinelli, M. Magliocchetti, C.J.A.P. Martins, Ana Paulino-Afonso, V. Capobianco, L. Moscardini, J Hjorth, S.-M Niemi, T. Schrabback, L. Gabarra, M Miluzio, I. A. Zinchenko, S. Mei, E. Zucca, S. Andreon, S. Maurogordato, S. V. H. Haugan, D. Vergani, R. Massey, T. Vassallo, S. Matthew, A Enia, Valérie Gautard, R. Toledo-Moreo, S Bruton, M. Moresco, W.G. Hartley, J. Skottfelt, B. R. Granett, Y Fang, D. Sapone, D. Maino, T. Castro, M. Schultheis, M. Meneghetti, H. J. McCracken, A. Fontana, A. Veropalumbo, L. Patrizii, G. Congedo, L. Valenziano, T.I Liaudat, H. Hildebrandt, S. Ferriol, M. Salvato, E. Romelli, C. Burigana, S Quai, Christopher J. Conselice, A. Biviano, I. Tutusaus, H. Degaudenzi, G. Zamorani, M Lembo, N. Martinet, L.A Popa, I Risso, V. Allevato, J. Väliviita, C. Neissner, D. Karagiannis, S. Escoffier, C. Padilla Aranda, L Blot, E. Maiorano, M Sahlén, M. Roncarelli, C. Kirkpatrick, R. B. Metcalf, A. Hornstrup, L. Conversi, S. Fotopoulou, E. Medinaceli, Sergey Kruk, M. Pöntinen, S. Di Domizio, Francesco Shankar, F. M. Zerbi, M. Béthermin, K. Jahnkę, A. Secroun, M. Farina, Antonello Calabrò, F. Mannucci, F. Pasian, V. Kansal, M. Castellano, F Lepori, F. Giacomini, F. Raison, M. Walmsley, J.A Schewtschenko, J Martín-Fleitas, Ariel G. Sánchez, M. Siudek, Q. Le Boulc'h, S. Dusini, N. A. Walton, B. Margalef-Bentabol, G Cañas-Herrera, Michael J. Smith, S. Davini, Antonio M. Mora, Alain Blanchard, I.T Andika, R. Cabanac, S. Camera, F. Hormuth, K Tanidis, L. Legrand, E. Bozzo, Romain Teyssier, LAURENCE C. SMITH, K. C. Chambers, M. Kilbinger, A. Costille, T Gasparetto, V. Lindholm, R. Rébolo, G. Verdoes Kleijn, P.-F Rocci, M. Sereno, M. Cropper, R. Nakajima, E. Keihänen, G. Meylan, G. Castignani, P. Hudelot, S. Cavuoti, F. Ricci, Petra Schneider, G Sirri, C. Carbone, Shoubaneh Hemmati, B. Altieri, K. Paterson, S. Kermiche, J. Rhodes, E Sarpa, Sune Toft, J. M. Diego, V. Pettorino, R. J. Bouwens, P. B. Lilje, S. Serrano, G. Polenta, M. Talia, A Pezzotta, A.M.C Le Brun, Aurel Schneider, D. Potter, E. Branchini, M. Kümmel, L. Pozzetti, O. Marggraf, S. Paltani, Y. Akrami, G. De Lucia, F. Atrio‐Barandela, H. M. Courtois, F. Cogato, D Sciotti, W. Roster, Yun Wang, A. Da Silva, L. Stančo, L. Maurin, C. M. Gutiérrez, J.E. Kajava, G. Testera, S Alvi, G Leroy, M. Tenti, I. Tereno, A. Zacchei, P Battaglia, M. Melchior, J. Macias-Perez, H. Kurki‐Suonio, M. Tucci, M. Wiesmann, A.N Taylor, K Naidoo, M. Poncet, J.A. Escartin Vigo, B. Kubik, I. Lloro, A Cimatti, G Verza, F Caro, F. Gianotti, N. Aghanim, J Lesgourgues, M. N. Bremer, A Finoguenov, M Archidiacono, C Colodro-Conde, M Guidi, G Morgante, X. Dupac, A. Renzi, T. Matamoro Zatarain, L. Leuzzi, Y Kang
article en

Abstract

Light emission from galaxies exhibit diverse brightness profiles, influenced by factors such as galaxy type, structural features, and interactions with other galaxies. Elliptical galaxies feature more uniform light distributions, while spiral and irregular galaxies have complex, varied light profiles due to their structural heterogeneity and star-forming activity. In addition, galaxies with active galactic nuclei (AGN) feature intense, concentrated emission from gas accretion around supermassive black holes, superimposed on regular galactic light, while quasi-stellar objects (QSOs) represent extreme cases in which AGN emissions dominate their host galaxies. The challenge of identifying AGN and QSOs has been discussed many times in the literature, often requiring multi-wavelength observations. This paper introduces a novel approach to identify AGN and QSOs from a single image. Diffusion models have recently been developed in the machine-learning literature to generate realistic-looking images of everyday objects. Utilising the spatial resolving power of the Euclid VIS images, we created a diffusion model trained on one million sources, without using any source pre-selection or labels. The model learns to reconstruct light distributions of normal galaxies, since the population is dominated by them. We conditioned the prediction of the central light distribution by masking the central few pixels of each source and reconstructed the light according to the diffusion model. We further used this prediction to identify sources that deviate from this profile by examining the reconstruction error of the few central pixels regenerated in each source's core. Our approach, solely using VIS imaging, features high completeness compared to traditional methods of AGN and QSO selection, including optical, near-infrared, mid-infrared, and X-rays. Our study offers practical insights for refining diffusion models and broadening their applications throughout the Euclid survey area, underscoring the utility of this approach in diverse astronomical contexts beyond just AGN identification.

Astronomy and Astrophysics
Openalex Percentile: Top 71%
Medical Image Segmentation Techniques
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.