Astronomers discover over 800 cosmic anomalies using a new AI tool
Here’s one use of AI that appears to do more good than harm. Two astronomers from the European Space Agency (ESA) have developed a neural network that searches for anomalies in space images. The results far exceeded what human experts could have done. In two and a half days, he sifted through nearly 100 million image cutouts, discovering 1,400 anomalous objects.
The creators of the AI model, David O’Ryan and Pablo Gómez, call it AnomalyMatch. The two men trained it (and applied it) to the Hubble Legacy Archive, which houses tens of thousands of datasets from Hubble’s 35-year history. “While trained scientists excel at detecting cosmic anomalies, there is simply too much Hubble data for experts to manually sort through to the necessary level of detail,” the ESA wrote in its press release.
After less than three days of analysis, AnomalyMatch returned a list of probable anomalies. It still requires human eyes in the end: Gómez and O’Ryan examined the candidates to confirm which ones were truly abnormal. Among the 1,400 anomalous objects confirmed by the two researchers, more than 800 were previously undocumented.
Most of the results showed galaxies merging or interacting, which can lead to strange shapes or long tails of stars and gas. Others were gravitational lenses. (This is where the gravity of a foreground galaxy bends space-time so that light from a background galaxy distorts into a circle or arc.) Other discoveries include planet-forming disks seen head-on, galaxies with huge star clusters, and jellyfish galaxies. Adding a bit of mystery, there were even “several dozen objects that defied classification.”
“This is a fantastic use of AI to maximize the scientific output of the Hubble archive,” Gómez said in the ESA statement. “Finding so many anomalous objects in the Hubble data, where one would expect many of them to have already been found, is a great result. It also shows how useful this tool will be for other large datasets.”
