A computer model of the James Webb Space Telescope’s infrared camera has helped researchers reveal an exoplanet that was hidden in existing observations, according to a ScienceDaily report published November 10, 2026. The approach, called Physics-based Learning for Imaging Planets (FLIP), also showed two to 15 times greater detection sensitivity than conventional methods in tests involving known planets, the report says.

The result matters because directly imaging a planet near a bright star is difficult: the star’s light can overwhelm the much fainter planet. FLIP aims to separate those signals by modeling how Webb’s Near Infrared Camera (NIRCam) produces an image. Researchers then use the model to estimate the star’s glare and subtract it from observations already collected by the telescope.

The technique uses a “digital twin”—a virtual replica of the camera—to account for subtle imperfections and changes in the instrument. The report says the researchers examined millions of possible variations that could affect the observations, refining the model until it more closely reproduced what the camera recorded. The method combines physical calculations of light and telescope optics with optimization techniques used in machine learning.

That distinction is central to the result: FLIP processes existing images; it does not make the telescope collect new light or demonstrate that every faint object revealed by analysis is a planet. The report describes one previously hidden exoplanet in Webb data and improved sensitivity in tests of known planets. It does not name the recovered planet or its host star, and it gives no detailed test results beyond the sensitivity range.

The work was reported in a Nature paper titled “Imaging exoplanets below the telescope stability limit,” published online November 4, 2026. The paper lists Rodrigo Ferrer-Chavez as lead author and Jason J. Wang and Katherine L. Bouman among its coauthors, along with William T. Freeman, Laurent Pueyo, Brandon Y. Feng and Marshall D. Perrin. ScienceDaily says its report was based on materials provided by the California Institute of Technology and was originally written by Katie Neith.

A potential next use is to revisit Webb’s archive. If FLIP can be applied successfully to more observations, images previously judged to show only a bright star could contain recoverable signals from nearby planets. The researchers also plan to investigate its use with other instruments, including the Nancy Grace Roman Space Telescope and ground-based telescopes. Those are future possibilities, not results established by the reported demonstration.

The longer-term motivation is to improve direct imaging of planets, including worlds in the habitable zones of Sun-like stars. The report notes that such targets would be extremely challenging because an Earth-like planet can be about 10 billion times fainter than its star. FLIP could help future missions by accounting computationally for some instrument imperfections after image capture, potentially easing the burden on telescope hardware. The work does not show that the technique can detect life or that it has yet achieved those future mission goals.