"Treasure Trove of Data: Scientists Unleash AI on Hubble's Archives to Discover Hidden Anomalies"
A team of astronomers from the European Space Agency (ESA) has made a groundbreaking discovery by utilizing an AI model to sift through NASA's 35-year-old Hubble dataset, uncovering over 800 previously unknown "astrophysical anomalies." The anomalies, which include weirdly shaped galaxies, light warped by massive objects, and planet-forming discs seen edge-on, were flagged by the AI model AnomalyMatch in just two and a half days.
The researchers, David O'Ryan and Pablo Gómez, trained the AI to scour nearly 100 million image cutouts from the Hubble Legacy Archive, which has never been systematically searched for anomalies before. The results, published in the journal Astronomy & Astrophysics, revealed nearly 1,400 anomalous objects, most of which were galaxies merging or interacting.
The discovery is a significant breakthrough in the field of astrophysics, where space is often found to be weird and unpredictable. The use of AI in this context is a testament to its ability to sift through vast amounts of information to spot patterns that may have gone unnoticed by human researchers.
"This is a fantastic use of AI to maximise the scientific output of the Hubble archive," said Gómez. "Finding so many anomalous objects in Hubble data, where you might expect many to have already been found, is a great result."
The ESA has hailed the discovery as a significant breakthrough and believes that the tool will be invaluable for other large datasets. As one researcher noted, "There's lots of it [data]...it's noisy, and the flood of data generated by tools like the Hubble Space Telescope can overwhelm even large research teams."
A team of astronomers from the European Space Agency (ESA) has made a groundbreaking discovery by utilizing an AI model to sift through NASA's 35-year-old Hubble dataset, uncovering over 800 previously unknown "astrophysical anomalies." The anomalies, which include weirdly shaped galaxies, light warped by massive objects, and planet-forming discs seen edge-on, were flagged by the AI model AnomalyMatch in just two and a half days.
The researchers, David O'Ryan and Pablo Gómez, trained the AI to scour nearly 100 million image cutouts from the Hubble Legacy Archive, which has never been systematically searched for anomalies before. The results, published in the journal Astronomy & Astrophysics, revealed nearly 1,400 anomalous objects, most of which were galaxies merging or interacting.
The discovery is a significant breakthrough in the field of astrophysics, where space is often found to be weird and unpredictable. The use of AI in this context is a testament to its ability to sift through vast amounts of information to spot patterns that may have gone unnoticed by human researchers.
"This is a fantastic use of AI to maximise the scientific output of the Hubble archive," said Gómez. "Finding so many anomalous objects in Hubble data, where you might expect many to have already been found, is a great result."
The ESA has hailed the discovery as a significant breakthrough and believes that the tool will be invaluable for other large datasets. As one researcher noted, "There's lots of it [data]...it's noisy, and the flood of data generated by tools like the Hubble Space Telescope can overwhelm even large research teams."