Galaxy morphology prediction using Capsule Networks

Reza Katebi*, Yadi Zhou, Ryan Chornock, Razvan Bunescu

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

3 Scopus citations


Understanding morphological types of galaxies is a key parameter for studying their formation and evolution. Neural networks that have been used previously for galaxy morphology classification have some disadvantages, such as not being inherently invariant under rotation. In this work, we studied the performance of Capsule Network (CapsNet), a recently introduced neural network architecture that is rotationally invariant and spatially aware, on the task of galaxy morphology classification.We designed two evaluation scenarios based on the answers from the question tree in the Galaxy Zoo project. In the first scenario, we used CapsNet for regression and predicted probabilities for all of the questions. In the second scenario, we chose the answer to the first morphology question that had the highest user agreement as the class of the object and trained a CapsNet classifier, where we also reconstructed galaxy images.We achieved promising results in both of these scenarios. Automated approaches such as the one introduced here will play a critical role in the upcoming large sky surveys.

Original languageEnglish (US)
Pages (from-to)1539-1547
Number of pages9
JournalMonthly Notices of the Royal Astronomical Society
Issue number2
StatePublished - Jun 1 2019
Externally publishedYes


  • catalogues
  • galaxy: general.
  • methods: data analysis
  • techniques: image processing

ASJC Scopus subject areas

  • Astronomy and Astrophysics
  • Space and Planetary Science

Fingerprint Dive into the research topics of 'Galaxy morphology prediction using Capsule Networks'. Together they form a unique fingerprint.

Cite this