Rotation-invariant convolutional neural networks for galaxy morphology prediction

2015/03/24 by Sander Dieleman, Kyle W. Willett, Kyle Willett +1 · 21 citations
Computer Science · Environmental Science · Mathematics · Physics and Astronomy · #Advanced Vision and Imaging #Archaeology #Artificial intelligence #Astrophysics #Computer science #Context (archaeology) #Convolutional neural network #Data Visualization and Analytics #Galaxy #Geography #Physics #Remote Sensing in Agriculture #astro-ph.GA #astro-ph.IM #cs.CV #cs.LG #cs.NE #stat.ML

paper · pdf · doi:10.1093/mnras/stv632

Accepted for publication in MNRAS. 20 pages, 14 figures

arxiv created 2015/03/24 · arxiv updated 2015/03/25 · openalex publication_date 2015/04/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05

Abstract

Measuring the morphological parameters of galaxies is a key requirement for studying their formation and evolution. Surveys such as the Sloan Digital Sky Survey have resulted in the availability of very large collections of images, which have permitted population-wide analyses of galaxy morphology. Morphological analysis has traditionally been carried out mostly via visual inspection by trained experts, which is time consuming and does not scale to large (≳104) numbers of images. Although attempts have been made to build automated classification systems, these have not been able to achieve the desired level of accuracy. The Galaxy Zoo project successfully applied a crowdsourcing strategy, inviting online users to classify images by answering a series of questions. Unfortunately, even this approach does not scale well enough to keep up with the increasing availability of galaxy images. We present a deep neural network model for galaxy morphology classification which exploits translational and rotational symmetry. It was developed in the context of the Galaxy Challenge, an international competition to build the best model for morphology classification based on annotated images from the Galaxy Zoo project. For images with high agreement among the Galaxy Zoo participants, our model is able to reproduce their consensus with near-perfect accuracy (>99 per cent) for most questions. Confident model predictions are highly accurate, which makes the model suitable for filtering large collections of images and forwarding challenging images to experts for manual annotation. This approach greatly reduces the experts’ workload without affecting accuracy. The application of these algorithms to larger sets of training data will be critical for analysing results from future surveys such as the Large Synoptic Survey Telescope.

Citations

• Theano: new features and speed improvements
• Spectral Networks and Locally Connected Networks on Graphs
• A logical calculus of the ideas immanent in nervous activity
• Galaxy types in the Sloan Digital Sky Survey using supervised artificial neural networks
• Automatic morphological classification of galaxy images
• Gradient-based learning applied to document recognition
• Galaxy Zoo: ‘Hanny's Voorwerp’, a quasar light echo?
• A Fast Learning Algorithm for Deep Belief Nets
• ANNz: Estimating Photometric Redshifts Using Artificial Neural Networks

Cited by

• Interpreting "Interpretability" and Explaining "Explainability" in Machine Learning in Physics
• The Evryscope Fast Transient Engine: Real-time Detection for Rapidly Evolving Transients
• Retinal vessel segmentation based on Fully Convolutional Neural Networks
• Deep learning for galaxy surface brightness profile fitting
• Machine learning and cosmological simulations – I. Semi-analytical models
• Fanaroff–Riley classification of radio galaxies using group-equivariant convolutional neural networks
• Observational Cosmology with Artificial Neural Networks
• Constraining the three-dimensional orbits of galaxies under ram pressure stripping with convolutional neural networks
• Improving galaxy morphologies for SDSS with Deep Learning
• Using convolutional neural networks to predict galaxy metallicity from three-colour images
• Bayesian bulge–disc decomposition of galaxy images
• Twenty-First-Century Statistical and Computational Challenges in Astrophysics
• Symmetry reduction for deep reinforcement learning active control of chaotic spatiotemporal dynamics
• A machine learning approach to galaxy–LSS classification – I. Imprints on halo merger trees
• A Hybrid Deep Learning Approach to Cosmological Constraints from Galaxy Redshift Surveys
• Deep learning topological invariants of band insulators
• Improving the Generalizability of Convolutional Neural Network-Based Segmentation on CMR Images
• Galaxy Zoo: quantitative visual morphological classifications for 48 000 galaxies from CANDELS
• Recurrent Existence Determination Through Policy Optimization
• Tuning target selection algorithms to improve galaxy redshift estimates
• Euclid preparation. Estimating galaxy physical properties using CatBoost chained regressors with attention
• Deep learning Blazar classification based on multifrequency spectral energy distribution data