DeepRed: an architecture for redshift estimation

Abstract

Estimating redshift is a central task in astrophysics, but its measurement is costly and time-consuming. In addition, current image-based methods are often validated on homogeneous datasets. The development and comparison of networks able generalize across different morphologies, ranging from galaxies to gravitationally-lensed transients, and observational conditions, remain an open challenge. This work proposes DeepRed, a deep learning pipeline that demonstrates how modern computer vision architectures, including ResNet, EfficientNet, Swin Transformer, and MLP-Mixer, can estimate redshifts from images of galaxies, gravitational lenses, and gravitationally-lensed supernovae. We compare these architectures and their ensemble to both neural networks (A1, A3, NetZ, and PhotoZ) and a feature-based method (HOG+SVR) on simulated (DeepGraviLens) and real (KiDS, SDSS) datasets. Our approach achieves state-of-the-art results on all datasets. On DeepGraviLens, DeepRed achieves a significant improvement in the Normalized Mean Absolute Deviation compared to the best baseline (PhotoZ): 55% on DES-deep (using EfficientNet), 51% on DES-wide (Ensemble), 52% on DESI-DOT (Ensemble), and 46% on LSST-wide (Ensemble). On real observations from the KiDS survey, the pipeline outperforms the best baseline (NetZ), improving NMAD by 16% on a general test set without high-probability lenses (Ensemble) and 27% on high-probability lenses (Ensemble). For non-lensed galaxies in the SDSS dataset, the MLP-Mixer architecture achieves a 5% improvement over the best baselines (A3 and NetZ). SHAP shows that the models correctly focus on the objects of interest with over 95% localization accuracy on high-quality images, validating the reliability of the predictions. These findings suggest that deep learning is a scalable, robust, and interpretable solution for redshift estimation in large-scale surveys.

Accepted for publication in Neural Computing and Applications

References in corpus (35)

• The Sloan Digital Sky Survey: Technical Summary
• The Dark Energy Camera
• A Perspective on Explainable Artificial Intelligence Methods: SHAP and LIME
• A Comprehensive Analysis of Deep Regression
• The fourth data release of the Kilo-Degree Survey: ugri imaging and nine-band optical-IR photometry over 1000 square degrees
• First Cosmology Results using Type Ia Supernovae from the Dark Energy Survey: Constraints on Cosmological Parameters
• Distance measures in cosmology
• ANNz2 - photometric redshift and probability distribution function estimation using machine learning
• Photometric redshifts from SDSS images using a Convolutional Neural Network
• Photometric redshift estimation via deep learning
• Galaxy And Mass Assembly (GAMA): autoz spectral redshift measurements, confidence and errors
• A Hybrid Ensemble method for Pulsar Candidate Classification
• Big Universe, Big Data: Machine Learning and Image Analysis for Astronomy
• The LSST Data Management System
• Tuning the Legacy Survey of Space and Time (LSST) Observing Strategy for Solar System Science
• Classification of Supernovae
• On the compared accuracy and reliability of spectroscopic and photometric redshift measurements
• ExoMiner: A Highly Accurate and Explainable Deep Learning Classifier that Validates 301 New Exoplanets
• Photometric Redshift Estimation with a Convolutional Neural Network: NetZ
• Three Gravitationally Lensed Supernovae Behind CLASH Galaxy Clusters
• The kinematic origin of the cosmological redshift
• The miniJPAS survey: star-galaxy classification using machine learning
• Investigating Deep Learning Methods for Obtaining Photometric Redshift Estimations from Images
• The first large catalogue of spectroscopic redshifts in Webb's First Deep Field, SMACS J0723.3$-$7327
• Photometric Redshifts from SDSS Images with an Interpretable Deep Capsule Network
• Optical and ultraviolet pulsed emission from an accreting millisecond pulsar
• On the gravitational redshift
• Wave-optical study of the Einstein cross formed by a quadrupole gravitational lens
• DeepZipper: A Novel Deep Learning Architecture for Lensed Supernovae Identification
• AI-driven spatio-temporal engine for finding gravitationally lensed type Ia supernovae
• TEGLIE: Transformer encoders as strong gravitational lens finders in KiDS
• Machine learning synthetic spectra for probabilistic redshift estimation: SYTH-Z
• A new pulsating neutron star in the Ultraluminous X-ray source NGC 4559 X7?
• KiDS-Legacy: Redshift distributions and their calibration
• DeepGraviLens: a Multi-Modal Architecture for Classifying Gravitational Lensing Data