DiffSketchEdit: Text-based Vector Sketch Editing

DiffSketchEdit enables text-guided editing of vector sketches using diffusion models. It supports various editing operations including word replacement, prompt refinement, and attention reweighting.

Usage

import requests

API_URL = "https://api-inference.huggingface.co/models/jree423/diffsketchedit"
headers = {"Authorization": "Bearer YOUR_HF_TOKEN"}

# Word replacement example
payload = {
    "inputs": {
        "prompts": [
            "A painting of a squirrel eating a burger",
            "A painting of a rabbit eating a burger"
        ],
        "edit_type": "replace"
    }
}

response = requests.post(API_URL, headers=headers, json=payload)

Edit Types

• replace: Word swap editing
• refine: Prompt refinement
• reweight: Attention reweighting
• generate: Simple generation

License

MIT License