ARCHIVE QUERY

Semantic Search

Describe what you're looking for, in plain English — EARTHZI matches your words against every SAR and optical patch already in the archive.

i Archive search, not prediction — how this works

WHAT THIS DOES — AND DOESN'T DO

This is archive search, not prediction. RemoteCLIP's text encoder shares an embedding space with its image encoder, so a query like "flooded agricultural land" is compared against the gallery's existing images by visual-semantic similarity — the same lookup image-to-image retrieval does, just without a source patch in hand.

It does not detect flood risk, classify land as flood-prone, or predict anything about a location. It only finds images already in the gallery that visually resemble your description.

TRY:

⌕

Type a description to begin

Try something like "flooded agricultural land" or pick one of the example queries above.