ESA is looking for volunteers to classify galaxies

The European Space Agency (ESA) has launched a new citizen science project in which volunteers can classify the shapes of galaxies in images taken by the Euclid space telescope. Volunteers will be given the chance to view unpublished images of the telescope and possibly previously undiscovered galaxies. By Friday afternoon, over 114,000 volunteers had already taken part in the project and made over 124,000 classifications.

Euclid, which was launched in July 2023 and has been carrying out routine scientific observations since February 2024, is expected to send around 100 GB of data to Earth every day for the next six years. This enormous amount of data makes manual classification by scientists alone almost impossible, writes the ESA.

Answers train an AI

Volunteers are shown images of galaxies and can answer simple questions about their shape or signs of spiral arms. The results of the citizen science project will also be used to train an AI algorithm called ZooBot, which will help to manage large amounts of data in the future.

ZooBot will initially classify the "simpler" galaxies and submit more complex cases to human users for evaluation. After training by volunteers, ZooBot should be able to make detailed classifications for hundreds of millions of galaxies, creating the largest scientific catalog of its kind to date. The first data sets with tens of thousands of galaxies, selected from more than 800,000 images, are already available on the platform.

Researching the influence of dark matter

The Euclid project aims to investigate the influence of dark matter and dark energy on the visible universe. To this end, the telescope will study the shapes, distances and movements of billions of galaxies up to a distance of 10 billion light years. The centerpiece is a high-resolution telescope equipped with two cameras - one for the visible wavelength range and one for the near-infrared spectrum.

(mki)