Attention-gating for improved radio galaxy classification

Abstract

In this work we introduce attention as a state of the art mechanism for classification of radio galaxies using convolutional neural networks. We present an attention-based model that performs on par with previous classifiers while using more than 50% fewer parameters than the next smallest classic CNN application in this field. We demonstrate quantitatively how the selection of normalisation and aggregation methods used in attention-gating can affect the output of individual models, and show that the resulting attention maps can be used to interpret the classification choices made by the model. We observe that the salient regions identified by the our model align well with the regions an expert human classifier would attend to make equivalent classifications. We show that while the selection of normalisation and aggregation may only minimally affect the performance of individual models, it can significantly affect the interpretability of the respective attention maps and by selecting a model which aligns well with how astronomers classify radio sources by eye, a user can employ the model in a more effective manner.

18 pages, 16 figures, Published in MNRAS

References in corpus (43)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• The Gaia mission
• LOFAR: The LOw-Frequency ARray
• Revisiting the Unified Model of Active Galactic Nuclei
• Multiple Object Recognition with Visual Attention
• Attention in Natural Language Processing
• The LOFAR Two-metre Sky Survey - II. First data release
• Science with ASKAP - the Australian Square Kilometre Array Pathfinder
• The Coma cluster magnetic field from Faraday rotation measures
• The Fifteenth Data Release of the Sloan Digital Sky Surveys: First Release of MaNGA Derived Quantities, Data Visualization Tools and Stellar Library
• Radio galaxies with a `double-double' morphology: I - Analysis of the radio properties and evidence for interrupted activity in active galactic nuclei
• Radio galaxies and feedback from AGN jets
• Revisiting the Fanaroff-Riley dichotomy and radio-galaxy morphology with the LOFAR Two-Metre Sky Survey (LoTSS)
• A pilot study of the radio-emitting AGN population: the emerging new class of FR0 radio-galaxies
• A simulation-based analytic model of radio galaxies
• Classifying Radio Galaxies with Convolutional Neural Network
• The cosmic evolution of radio-AGN feedback to z=1
• Neural computation as a tool for galaxy classification : methods and examples
• Radio Galaxy Zoo: ClaRAN - A Deep Learning Classifier for Radio Morphologies
• FRICAT: A FIRST catalog of FRI radio galaxies
• Rotation Measures of Radio Sources in Hot Galaxy Clusters
• The nuclear properties and extended morphologies of powerful radio galaxies: the roles of host galaxy and environment
• The link between accretion mode and environment in radio-loud active galaxies
• Identifying galaxies, quasars, and stars with machine learning: A new catalogue of classifications for 111 million SDSS sources without spectra
• LoTSS DR1: Double-double radio galaxies in the HETDEX field
• High-resolution VLA observations of FR0 radio galaxies: properties and nature of compact radio sources
• The intergalactic magnetic field probed by a giant radio galaxy
• Morphological classification of radio galaxies: Capsule Networks versus Convolutional Neural Networks
• Cosmological Measurements with Forthcoming Radio Continuum Surveys
• A Machine Learning Based Morphological Classification of 14,245 Radio AGNs Selected From The Best-Heckman Sample
• MADE: A spectroscopic Mass, Age, and Distance Estimator for red giant stars with Bayesian machine learning
• On the population of remnant FRII radio galaxies and implications for radio source dynamics
• Radio Galaxy Zoo: A Search for Hybrid Morphology Radio Galaxies
• The CoNFIG Catalogue - II. Comparison of Space Densities in the FR Dichotomy
• X-ray study of a sample of FR0 radio galaxies: unveiling the nature of the central engine
• Support Vector Machine classification of strong gravitational lenses
• Science with the Murchison Widefield Array: Phase I Results and Phase II Opportunities
• The LOFAR view of FR0 radio galaxies
• The new class of FR0 radio galaxies
• The Combined NVSS-FIRST Galaxies (CoNFIG) Sample - I. Sample Definition, Classification and Evolution
• Machine Learning Classification of Gaia Data Release 2
• PKS 2250$-$351: A Giant Radio Galaxy in Abell 3936
• The morphological classification of distant radio galaxies explored with three-dimensional simulations

Cited by in corpus (25)

• Fanaroff-Riley classification of radio galaxies using group-equivariant convolutional neural networks
• Machine-learning classification of astronomical sources: estimating F1-score in the absence of ground truth
• Euclid preparation: XXII. Selection of Quiescent Galaxies from Mock Photometry using Machine Learning
• Quantifying Uncertainty in Deep Learning Approaches to Radio Galaxy Classification
• A Multimodal Dataset and Benchmark for Radio Galaxy and Infrared Host Detection
• DECORAS: detection and characterization of radio-astronomical sources using deep learning
• Feature Guided Training and Rotational Standardisation for the Morphological Classification of Radio Galaxies
• Morphological Classification of Radio Galaxies with wGAN-supported Augmentation
• Improving machine learning-derived photometric redshifts and physical property estimates using unlabelled observations
• 3D Detection and Characterisation of ALMA Sources through Deep Learning
• Challenging interferometric imaging: Machine learning-based source localization from uv-plane observations
• Finding AGN remnant candidates based on radio morphology with machine learning
• Euclid preparation. LI. Forecasting the recovery of galaxy physical properties and their relations with template-fitting and machine-learning methods
• Identification of 4876 Bent-Tail Radio Galaxies in the FIRST Survey Using Deep Learning Combined with Visual Inspection
• Identifying type II quasars at intermediate redshift with few-shot learning photometric classification
• Deep learning-based astronomical multimodal data fusion: A comprehensive review
• Insights on Galaxy Evolution from Interpretable Sparse Feature Networks
• Euclid preparation. Estimating galaxy physical properties using CatBoost chained regressors with attention
• E(2)-Equivariant Features in Machine Learning for Morphological Classification of Radio Galaxies
• Radio Galaxy Zoo: Using semi-supervised learning to leverage large unlabelled data-sets for radio galaxy classification under data-set shift
• Radio Galaxy Zoo: Giant Radio Galaxy Classification using Multi-Domain Deep Learning
• Can semi-supervised learning reduce the amount of manual labelling required for effective radio galaxy morphology classification?
• Weight Pruning and Uncertainty in Radio Galaxy Classification
• E(2) Equivariant Self-Attention for Radio Astronomy
• Radio Galaxy Zoo: Morphological classification by Fanaroff-Riley designation using self-supervised pre-training