Blank Canvas Diffusion v1 Model Card

Blank Canvas Diffusion is a text-to-image model without Graphic Art Knowledge, i.e. trained on Blank Canvas dataset.

This model is first trained 41,400 steps at the resolution of 256, then 156,700 steps at the resolution of 512

from diffusers import DiffusionPipeline

pipe = DiffusionPipeline.from_pretrained("rhfeiyang/Blank-Canvas-Diffusion-v1").to("cuda")
images = pipe("The image depicts a picturesque small town by a river, featuring several docked boats. Surrounded by trees, the town is near a large body of water, highlighting its popularity for boating and water activities. The serene composition, with trees and boats, underscores the town's natural beauty and tranquil charm.", 
          num_inference_steps=50, guidance_scale=7.5,).images

Citation

@misc{ren2025optinartlearningart,
        title={Opt-In Art: Learning Art Styles Only from Few Examples}, 
        author={Hui Ren and Joanna Materzynska and Rohit Gandikota and David Bau and Antonio Torralba},
        year={2025},
        eprint={2412.00176},
        archivePrefix={arXiv},
        primaryClass={cs.CV},
        url={https://arxiv.org/abs/2412.00176}, 
}