🛰️ DINOv3 Satellite Vision: Interactive Patch Similarity

Click any region to visualize feature similarities across the image

Model Selection

Select a model (size/pretraining dataset)
0 1
Example Images
Pages:
Satellite-pretrained models are intended for: geographic patterns, land use classification. structural analysis, etc. Try comparing similarity maps for the same image created by the model pretrained on sat493m vs. the one on lvd1689m (general web).

Dynamic Resolution: Images are processed at up to 720px (longer side) while preserving aspect ratio. DINOv3's 3D RoPE embeddings handle variable sizes.