Boffins use machine-learning tricks to create a starlight dimmer switch

Scientists have developed a machine-learning technique called Physics-based Learning for Imaging Planets (FLIP) to remove starlight from telescope images, improving sensitivity to faint exoplanets by up to 15 times. Led by Northwestern University's Rodrigo Ferrer-Chavez, the team combined a physics-based telescope model with optimization techniques to subtract unwanted starlight.

This approach successfully produced the first image of the giant planet 51 Eridani b at an infrared wavelength of 4.3 micrometers, where conventional processing had previously failed. The FLIP framework demonstrated sensitivity improvements of two to 15 times over conventional techniques, potentially reducing the need for unprecedented stability requirements in ground-based and space-based observatories.

Brief written by urgent.news from The Register Science's own syndicated text. Machine-written — may contain errors; check the original before relying on it.