N2N4M

Removes noise from Mars CRISM shortwave-infrared hyperspectral data.

Field
Planetary science
Method
Deep learning
Data it takes
imagery, spectra
License
MIT

CRISM shortwave-infrared spectra carry noise that obscures the absorption features mineral identification depends on. N2N4M denoises them without needing clean reference spectra to train against.

The point is what becomes visible afterwards: weak absorptions that were buried in noise become measurable, which changes what can be identified from orbit.

Code and weights are both MIT.

Upstream project →

Catalog entry last checked 2026-08-19. All models →

Run N2N4M on your data

The lab runs this on its own compute, in a container, with the inputs and parameters recorded alongside the result. Initial scoping conversations are free.

Contact the Lab