An artificial intelligence (AI) programme used to recognize faces on Facebook can also identify galaxies in deep space, scientists said Wednesday. The AI bot named ClaRAN scans images taken by radio telescopes, said researchers from the International Centre for Radio Astronomy Research (ICRAR) in Australia.
Its job is to spot radio galaxies — galaxies that emit powerful radio jets from supermassive black holes at their centers, according to the research published in the journal Monthly Notices of the Royal Astronomical Society.
Black holes are found at the center of most, if not all, galaxies.
“These supermassive black holes occasionally burp out jets that can be seen with a radio telescope,” said Ivy Wong from The University of Western Australia node of the International Centre for Radio Astronomy Research (ICRAR).
“Over time, the Jets can stretch a long way from their host galaxies, making it difficult for traditional computer programmes to figure out where the galaxy is,” said Wong.
Wong said the programme was completely overhauled and trained to recognize galaxies instead of people.
She said the upcoming EMU survey using the Australian Square Kilometre Array Pathfinder (ASKAP) telescope is expected to observe up to 70 million galaxies across the history of the universe.
Wong said traditional computer algorithms are able to correctly identify 90 percent of the sources.
“That still leaves 10 percent or seven million ‘difficult’ galaxies that have to be eyeballed by a human due to the complexity of their extended structures,” Wong said.
“If ClaRAN reduces the number of sources that require visual classification down to one percent, this means more time for our citizen scientists to spend looking at new types of galaxies,” she said.
A highly-accurate catalog produced by Radio Galaxy Zoo volunteers was used to train ClaRAN how to spot where the jets originate.
Chen Wu, also from ICRAR, said ClaRAN is an example of a new paradigm called ‘programming 2.0’.
“All you do is set up a huge neural network, give it a tonne of data, and let it figure out how to adjust its internal connections in order to generate the expected outcome,” he said.