AI Gives James Webb Telescope Super Sharp Vision

From left to right: Images of the galaxy NGC 1068, Jupiter's moon Io and Wolf-Rayet star 137, before (top) and after (bottom) AI sharpening. (Image credit: Max Charles/University of Sydney)

The James Webb Space Telescope, humanity's most powerful eye on the cosmos, recently encountered a significant challenge: one of its specialized instruments was returning blurry images. This issue threatened its mission to find distant exoplanets within our own Milky Way galaxy. However, a team of Australian researchers developed a groundbreaking AI algorithm that has successfully corrected the telescope's vision, proving that complex problems in space can sometimes be solved with brilliant code from Earth.

The Instrument Behind the Issue

The problem originated with the Aperture Masking Interferometer (API), a specialized device installed on one of the James Webb Space Telescope's four main instruments, the Near-InfraRed Imager and Slitless Spectrograph (NIRISS). Designed by a team led by Professor Peter Tuthill from the University of Sydney, the API uses a mask with seven holes to combine light from different parts of the telescope's primary mirror. This technique dramatically increases the instrument's resolution, making it perfect for spotting small, dim exoplanets orbiting distant stars. When the first images from the API came back blurry, it was traced to electronic distortions on Webb's infrared camera detector.

Echoes of Hubble But a Modern Solution

The situation was reminiscent of the famous flaw in the Hubble Space Telescope's primary mirror, which required a crewed shuttle mission in 1993 to install corrective optics. Such a physical repair is impossible for Webb, which operates nearly a million miles from Earth—more than three times the distance to the moon and far beyond the reach of any human mission. The solution had to be digital.

Enter AMIGO The AI Image Corrector

To tackle the problem, former University of Sydney Ph.D. students Max Charles and Louis Desdoigts created a sophisticated neural network called AMIGO (Aperture Masking Interferometry Generative Observations). This AI algorithm was trained to identify and correct the specific pixels affected by the electrical distortions, effectively sharpening the blurry images with software.

"Instead of sending astronauts to bolt on new parts, they managed to fix things with code," Professor Tuthill noted in a statement.

Sharpening Our View of the Cosmos

The results have been remarkable. The researchers demonstrated AMIGO's power by clarifying images of a dim exoplanet and a cool red-brown dwarf star located 133 light-years away. With its vision corrected, the API has also produced stunningly detailed images of a black hole jet, the volcanic surface of Jupiter's moon Io, and stellar winds from a distant variable star.

"This work brings JWST's vision into even sharper focus," said Desdoigts, who is now a postdoctoral researcher at Leiden University. "It's incredibly rewarding to see a software solution extend the telescope's scientific reach.”

Since becoming operational in 2022, the James Webb Space Telescope has already revolutionized our understanding of the universe. With the API now performing at its full potential thanks to AMIGO, Webb is poised to deliver even more breathtaking discoveries.