---
license: apache-2.0
tags:
  - controlnet
  - stable-diffusion
  - satellite-imagery
  - osm
  - image-to-image
  - diffusers
base_model: stabilityai/stable-diffusion-2-1-base
pipeline_tag: image-to-image
library_name: diffusers
---

# VectorSynth

**VectorSynth** is a ControlNet model that generates satellite imagery from OpenStreetMap (OSM) vector data embeddings. It conditions [Stable Diffusion 2.1 Base](https://huggingface.co/stabilityai/stable-diffusion-2-1-base) on rendered OSM text to synthesize realistic aerial imagery.

## Model Description

VectorSynth uses a two-stage pipeline:
1. **RenderEncoder**: Projects 768-dim CLIP text embeddings of OSM text to 3-channel control images
2. **ControlNet**: Conditions Stable Diffusion 2.1 on the rendered control images

This model uses standard CLIP embeddings. For the COSA embedding variant, see [VectorSynth-COSA](https://huggingface.co/MVRL/VectorSynth-COSA).

## Files

- `config.json` - ControlNet configuration
- `diffusion_pytorch_model.safetensors` - ControlNet weights
- `render_encoder/clip-render_encoder.pth` - RenderEncoder weights
- `render.py` - RenderEncoder class definition

## Citation

```bibtex
@inproceedings{cher2025vectorsynth,
  title={VectorSynth: Fine-Grained Satellite Image Synthesis with Structured Semantics},
  author={Cher, Daniel and Wei, Brian and Sastry, Srikumar and Jacobs, Nathan},
  year={2025},
  eprint={arXiv:2511.07744},
  note={arXiv preprint}
}
```

## Related Models

- [VectorSynth-COSA](https://huggingface.co/MVRL/VectorSynth-COSA) - COSA embedding variant
- [GeoSynth](https://huggingface.co/MVRL/GeoSynth) - Text-to-satellite image generation