As machine learning approaches become more sophisticated, they are increasingly used in astronomy for difficult tasks such as finding faint and distant galaxy clusters. Having computers sift through astronomical data to look for specific objects can be tremendously helpful, as they can handle a massive amount of data — however, there are some judgments that still require the human hand.
This week’s image from the Hubble Space Telescope shows an object discovered by a human despite being missed by a computer algorithm. The dwarf galaxy Donatiello II is very faint and difficult to see from the background beyond, but an amateur astronomer was able to spot it.
“Even the best algorithms have limitations when it comes to distinguishing very faint galaxies from single stars and background noise,” write Hubble scientists. “In such challenging situations, identification needs to be done the old-fashioned way — by a dedicated human digging through the data themselves.”
Located at the center of this image, the dwarf galaxy was identified by Giuseppe Donatiello from data collected during the Dark Energy Survey. It orbits the Sculptor Galaxy along with two other similar dwarf galaxies named Donatiello III and IV. When Donatiello spotted the trio in the Dark Energy Survey data, researchers used Hubble to confirm the discovery and capture this image.
The Dark Energy Survey was a project that surveyed the sky from 2013 to 2019, examining objects such as galaxy clusters to learn more about dark energy. Although the survey is now complete, the data collected over the first three years of observing was released to the public in 2021 and is still leading to discoveries like this trio of dwarf galaxies.