AI System Analyzes Hubble Images, Identifies Over 1,300 Cosmic Oddities
A machine-learning system named AnomalyMatch has processed nearly 100 million image cutouts from the Hubble Legacy Archive, identifying 1,339 unusual cosmic sources. This effort, led by David O’Ryan and Pablo Gómez of the European Space Agency, resulted in a catalogue where 811 of these sources had no prior reference in scientific literature. The AI system did not autonomously declare discoveries but ranked the images for astronomers to review. The findings include rare versions of known processes such as galaxy mergers and gravitational lenses. The study emphasizes that these anomalies are not necessarily new phenomena but rather objects with unique visible shapes within the dataset. 
