DemoFusion
Code release for "DemoFusion: Democratising High-Resolution Image Generation With No 💰" (arXiv）
Abstract: High-resolution image generation with Generative Artificial Intelligence (GenAI) has immense potential but, due to the enormous capital investment required for training, it is increasingly centralised to a few large corporations, and hidden behind paywalls. This paper aims to democratise high-resolution GenAI by advancing the frontier of high-resolution generation while remaining accessible to a broad audience. We demonstrate that existing Latent Diffusion Models (LDMs) possess untapped potential for higher-resolution image generation. Our novel DemoFusion framework seamlessly extends open-source GenAI models, employing Progressive Upscaling, Skip Residual, and Dilated Sampling mechanisms to achieve higher-resolution image generation. The progressive nature of DemoFusion requires more passes, but the intermediate results can serve as "previews", facilitating rapid prompt iteration.
News
- 2023.11.29: 'pipeline_demofusion_sdxl' is released.
Usage
- The version requirements of core dependencies.
torch==2.1.0
diffusers==0.21.4
- Download
pipeline_demofusion_sdxl.pyand run it as follows.
from pipeline_demofusion_sdxl import DemoFusionSDXLPipeline
model_ckpt = "stabilityai/stable-diffusion-xl-base-1.0"
pipe = DemoFusionSDXLPipeline.from_pretrained(model_ckpt, torch_dtype=torch.float16)
pipe = pipe.to("cuda")
prompt = "Envision a portrait of an elderly woman, her face a canvas of time, framed by a headscarf with muted tones of rust and cream. Her eyes, blue like faded denim. Her attire, simple yet dignified."
negative_prompt = "blurry, ugly, duplicate, poorly drawn, deformed, mosaic"
images = pipe(prompt, negative_prompt=negative_prompt,
height=3072, width=3072, view_batch_size=16, stride=64,
num_inference_steps=50, guidance_scale=7.5,
cosine_scale_1=3, cosine_scale_2=1, cosine_scale_3=1, sigma=0.8,
multi_decoder=True, show_image=True
)
- Please feel free to try different prompts and resolutions.
- Default hyper-parameters are recommended, but they may not be optimal for all cases. For specific impacts of each hyper-parameter, please refer to Appendix C in the DemoFusion paper.
- The code was cleaned before the release. If you encounter any issues, please contact us.
- A use case can be found in
demo.ipynb.
Citation
If you find this paper useful in your research, please consider citing:
@article{du2023demofusion,
title={DemoFusion: Democratising High-Resolution Image Generation With No $$$},
author={Ruoyi Du and Dongliang Chang and Timothy M. Hospedales and Yi-Zhe Song and Zhanyu Ma},
journal={arXiv},
year={2023}
}