tangkai-RS/AnytimeFormer

[RSE 2026] AnytimeFormer: Fusing irregular and asynchronous SAR-optical time series to reconstruct reflectance at any given time

28
/ 100
Experimental

AnytimeFormer helps remote sensing professionals reconstruct clear images of Earth's surface reflectance, even when data is sparse or irregular. It takes radar (SAR) and optical satellite imagery collected at different times and stitches them together. The output is a complete, gap-free time series of surface reflectance, providing a consistent view of changes over time for environmental monitoring or land use analysis.

Use this if you need to generate a continuous, high-quality time series of surface reflectance from patchy or inconsistent SAR and optical satellite data.

Not ideal if you primarily work with perfect, complete satellite image datasets and do not need to fill in significant data gaps caused by clouds or sensor limitations.

remote-sensing earth-observation environmental-monitoring land-use-analysis satellite-image-fusion
No License No Package No Dependents
Maintenance 6 / 25
Adoption 7 / 25
Maturity 7 / 25
Community 8 / 25

How are scores calculated?

Stars

39

Forks

3

Language

Jupyter Notebook

License

Last pushed

Jan 13, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/tangkai-RS/AnytimeFormer"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.