Actions
ecbyts
ecbyts
ecbyts
Adiciona assinatura digital, hash-chain e Merkle Tree para verificacao de integridade e autoria.
Vincula este modelo a uma carteira Ethereum. Modo privado: registro no Supabase. Modo público: NFT na Base L2.
This key contains all your model data. Share it or save it to restore later.
ollama pull gemma4:e2b
(downloads ~7 GB, only once)
Click on the image to place annotations
Click on the image to segment objects. Left-click = include, Right-click = exclude.
SlimSAM: optimized for browser, single-object focus
Automatically detects all objects in the image. No clicks needed.
Describe what to detect. CLIP classifies auto-generated masks by text similarity.
CLIP+SAM models require ~135 MB download (cached after first use)
Select a map region to auto-segment satellite imagery (Sentinel-2).
Compare two images to detect environmental changes over time.
Current image = reference. Upload a second image for comparison.
Classify the scene type of the current image (forest, water, urban, etc.).
Compute vegetation indices (vNDVI, SAVI) from RGB imagery.
Index images for similarity search across tiles and time periods.
Regra de elevacao: UTM usa elevacao absoluta; X/Z usa elevacao relativa.
| Substance | CAS / ID | VR | VP | VI | Unit | Source |
|---|