Galaxy Morphology Classification with Deep Convolutional Neural Networks

TL;DR

This paper introduces a residual network variant for galaxy morphology classification, achieving state-of-the-art accuracy on Galaxy Zoo 2 data, and demonstrates its potential for large-scale astronomical surveys.

Contribution

A novel residual network variant that outperforms existing CNNs in galaxy classification tasks, with high accuracy and applicability to future large-scale surveys.

Findings

Achieved 95.21% overall accuracy on Galaxy Zoo 2 dataset.

Outperformed popular CNN architectures like AlexNet, VGG, Inception.

High classification accuracy across different galaxy types.

Abstract

We propose a variant of residual networks (ResNets) for galaxy morphology classification. The variant, together with other popular convolutional neural networks (CNNs), are applied to a sample of 28790 galaxy images from Galaxy Zoo 2 dataset, to classify galaxies into five classes, i.e. completely round smooth, in-between smooth (between completely round and cigar-shaped), cigar-shaped smooth, edge-on and spiral. A variety of metrics, such as accuracy, precision, recall, F1 value and AUC, show that the proposed network achieves the state-of-the-art classification performance among the networks, namely, Dieleman, AlexNet, VGG, Inception and ResNets. The overall classification accuracy of our network on the testing set is 95.2083% and the accuracy of each type is given as: completely round, 96.6785%; in-between, 94.4238%; cigar-shaped, 58.6207%; edge-on, 94.3590% and spiral, 97.6953% respectively. Our model algorithm can be applied to large-scale galaxy classification in forthcoming surveys such as the Large Synoptic Survey Telescope (LSST).