How to See the Invisible: Recovering Polarization from a Handful of Photons ⚡ экспресс
Scientists found a way to measure light polarization in ultra-low light conditions—when photons are so few that conventional methods fail. The new approach, akin to reconstructing a painting from a few brushstrokes, uses the relationships between adjacent data points and machine learning. This allows extracting detailed polarization information across wavelengths even in near-total darkness. Imagine: how can just a handful of photons reveal a material's hidden properties?
When light is scarce, a camera can't determine its polarization—the image turns into chaos of scattered dots. It's like trying to restore a painting from a few random brushstrokes. But if the strokes lie along a line, an experienced eye guesses what's missing.
Researchers have created an algorithm that uses the same principle. It knows that polarization (the orientation of a light wave) usually changes smoothly from one color to another—it was trained on examples.
So, from a handful of spectroscopic (color) data, it predicts missing values.
The method requires no complex equipment and works even with single photons. It will be useful for studying distant exoplanets—for example, to assess whether they have vegetation from reflected light, even if only a few light particles reach us. Or for ultra-sensitive photometry (brightness measurement) in laboratories.
🎯 Ordinary polarizing glasses are almost useless at twilight—they block some light, making an already dim picture even darker. The new method recovers light orientation from scarce data, restoring the ability to see in the dark.