lucataco/demofusion

DemoFusion: Democratising High-Resolution Image Generation With No 💰

  • Public
  • 7.8K runs

Run time and cost

This model runs on Nvidia A100 (80GB) GPU hardware. Predictions typically complete within 7 minutes. The predict time for this model varies significantly based on the inputs.

Readme

This is an implementation of DemoFusion. DemoFusion provides highres txt2img capabilities based on SDXL. See the demo example that showcases a txt to img run that provides an img in 1024x1024, 2048x2048, and 3076x3076 resolution in under 6min

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.

@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}
}