Practical Galaxy Morphology Tools from Deep Supervised Representation Learning

Abstract

Astronomers have typically set out to solve supervised machine learning problems by creating their own representations from scratch. We show that deep learning models trained to answer every Galaxy Zoo DECaLS question learn meaningful semantic representations of galaxies that are useful for new tasks on which the models were never trained. We exploit these representations to outperform several recent approaches at practical tasks crucial for investigating large galaxy samples. The first task is identifying galaxies of similar morphology to a query galaxy. Given a single galaxy assigned a free text tag by humans (e.g. "#diffuse"), we can find galaxies matching that tag for most tags. The second task is identifying the most interesting anomalies to a particular researcher. Our approach is 100% accurate at identifying the most interesting 100 anomalies (as judged by Galaxy Zoo 2 volunteers). The third task is adapting a model to solve a new task using only a small number of newly-labelled galaxies. Models fine-tuned from our representation are better able to identify ring galaxies than models fine-tuned from terrestrial images (ImageNet) or trained from scratch. We solve each task with very few new labels; either one (for the similarity search) or several hundred (for anomaly detection or fine-tuning). This challenges the longstanding view that deep supervised methods require new large labelled datasets for practical use in astronomy. To help the community benefit from our pretrained models, we release our fine-tuning code Zoobot. Zoobot is accessible to researchers with no prior experience in deep learning.

arXiv comment20 pages plus appendix. Accepted to MNRAS (open-access DOI below). Code, documentation, pretrained models: https://github.com/mwalmsley/zoobot (PyTorch and TensorFlow)

References in corpus (49)

• The Astropy Project: Building an inclusive, open-science project and status of the v2.0 core package
• scikit-image: Image processing in Python
• Overview of the DESI Legacy Imaging Surveys
• Rotation-invariant convolutional neural networks for galaxy morphology prediction
• Galaxy Zoo 2: detailed morphological classifications for 304,122 galaxies from the Sloan Digital Sky Survey
• Galaxy Zoo Green Peas: Discovery of A Class of Compact Extremely Star-Forming Galaxies
• The Thirteenth Data Release of the Sloan Digital Sky Survey: First Spectroscopic Data from the SDSS-IV Survey MApping Nearby Galaxies at Apache Point Observatory
• A Catalog of Detailed Visual Morphological Classifications for 14034 Galaxies in the Sloan Digital Sky Survey
• Improving galaxy morphologies for SDSS with Deep Learning
• Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity
• A Classical Morphological Analysis of Galaxies in the Spitzer Survey of Stellar Structure in Galaxies (S4G)
• Galaxy Zoo : 'Hanny's Voorwerp', a quasar light echo?
• Planet Hunters X. KIC 8462852 - Where's the Flux?
• Towards Fairer Datasets: Filtering and Balancing the Distribution of the People Subtree in the ImageNet Hierarchy
• KOI-54: The Kepler Discovery of Tidally-Excited Pulsations and Brightenings in a Highly Eccentric Binary
• Galaxy Zoo DECaLS: Detailed Visual Morphology Measurements from Volunteers and Deep Learning for 314,000 Galaxies
• Galaxy Zoo: Probabilistic Morphology through Bayesian CNNs and Active Learning
• Using transfer learning to detect galaxy mergers
• Rings and spirals in barred galaxies. I Building blocks
• Galaxy Zoo: comparing the demographics of spiral arm number and a new method for correcting redshift bias
• The weirdest SDSS galaxies: results from an outlier detection algorithm
• ARRAKIS: Atlas of Resonance Rings As Known In the S4G
• SDSS-IV MaNGA PyMorph Photometric and Deep Learning Morphological Catalogs and implications for bulge properties and stellar angular momentum
• Radio Galaxy Zoo: ClaRAN - A Deep Learning Classifier for Radio Morphologies
• Machine and Deep Learning Applied to Galaxy Morphology -- A Comparative Study
• Transfer learning for galaxy morphology from one survey to another
• An automatic taxonomy of galaxy morphology using unsupervised machine learning
• Astronomaly: Personalised Active Anomaly Detection in Astronomical Data
• A new catalogue of polar-ring galaxies selected from the SDSS
• Self-Supervised Representation Learning for Astronomical Images
• Galaxy Zoo: Bulgeless Galaxies With Growing Black Holes
• Identifying galaxies, quasars, and stars with machine learning: A new catalogue of classifications for 111 million SDSS sources without spectra
• The relationship between fine galaxy stellar morphology and star formation activity in cosmological simulations: a deep learning view
• Beyond the Hubble Sequence -- Exploring Galaxy Morphology with Unsupervised Machine Learning
• Galaxy Morphology Network: A Convolutional Neural Network Used to Study Morphology and Quenching in $\sim 100,000$ SDSS and $\sim 20,000$ CANDELS Galaxies
• Radio Galaxy Zoo: Unsupervised Clustering of Convolutionally Auto-encoded Radio-astronomical Images
• Probing the Evolution of the Galaxy Interaction/Merger Rate Using Collisional Ring Galaxies
• AstroVaDEr: Astronomical Variational Deep Embedder for Unsupervised Morphological Classification of Galaxies and Synthetic Image Generation
• Galaxy Zoo: The interplay of quenching mechanisms in the group environment
• Galactic Rings Revisited. I. CVRHS Classifications of 3962 Ringed Galaxies from the Galaxy Zoo 2 Database
• Transfer learning for radio galaxy classification
• Anomaly detection in Hyper Suprime-Cam galaxy images with generative adversarial networks
• A Comprehensive Examination of the Optical Morphologies of 719 Isolated Galaxies in the AMIGA Sample
• Applying the Analytic Theory of Colliding Ring Galaxies
• Capturing the physics of MaNGA galaxies with self-supervised Machine Learning
• Multiband galaxy morphologies for CLASH: a convolutional neural network transferred from CANDELS
• Automatic detection of full ring galaxy candidates in SDSS
• Morphology-assisted galaxy mass-to-light predictions using deep learning
• Survey2Survey: A deep learning generative model approach for cross-survey image mapping

Cited by in corpus (4)

• A unique, ring-like radio source with quadrilateral structure detected with machine learning
• Machine learning technique for morphological classification of galaxies from the SDSS. III. Image-based inference of detailed features
• Analysis of Ring Galaxies Detected Using Deep Learning with Real and Simulated Data
• Learning to Detect Interesting Anomalies