Home

Awesome

<div align="center"> <h2>Make a Cheap Scaling: A Self-Cascade Diffusion Model for Higher-Resolution Adaptation</h2>

<a href='https://arxiv.org/abs/2402.10491'><img src='https://img.shields.io/badge/ArXiv-2305.18247-red'></a>      <a href='https://guolanqing.github.io/Self-Cascade/'><img src='https://img.shields.io/badge/Project-Page-Green'></a>

Lanqing Guo*, Yingqing He*, Haoxin Chen, Menghan Xia, Xiaodong Cun, Yufei Wang, Siyu Huang,<br> Yong Zhang<sup>#, Xintao Wang, Qifeng Chen, Ying Shan and Bihan Wen<sup>#

(* first author, # corresponding author)

</div>

🥳 Demo

<p align="center"> <img src="assets/video_demo.gif" width="700px"> </p>

Please check more demo videos at the project page.

🔆 Abstract

<b>TL; DR: 🤗🤗🤗 Self-cascade diffusion model is a lightweight and efficient scale adaptation approach for higher-resolution image and video generation.</b>

Diffusion models have proven to be highly effective in image and video generation; however, they still face composition challenges when generating images of varying sizes due to single-scale training data. Adapting large pre-trained diffusion models for higher resolution demands substantial computational and optimization resources, yet achieving a generation capability comparable to low-resolution models remains elusive. This paper proposes a novel self-cascade diffusion model that leverages the rich knowledge gained from a well-trained low-resolution model for rapid adaptation to higher-resolution image and video generation, employing either tuning-free or cheap upsampler tuning paradigms. Integrating a sequence of multi-scale upsampler modules, the self-cascade diffusion model can efficiently adapt to a higher resolution, preserving the original composition and generation capabilities. We further propose a pivot-guided noise re-schedule strategy to speed up the inference process and improve local structural details. Compared to full fine-tuning, our approach achieves a 5X training speed-up and requires only an additional 0.002M tuning parameters. Extensive experiments demonstrate that our approach can quickly adapt to higher resolution image and video synthesis by fine-tuning for just 10k steps, with virtually no additional inference time.

🔥 Update

🔎 Main Requirements

This repository is tested on

💫 Inference

Text-to-image higher-resolution generation with diffusers script

stable-diffusion xl v1.0 base

# 2048x2048 (4x) generation
python3 sdxl_inference.py \
--validation_prompt "a professional photograph of an astronaut riding a horse" \
--seed 23 \
--mode tuning

💫 Tuning

😉 Citation

@article{guo2024make,
  title={Make a Cheap Scaling: A Self-Cascade Diffusion Model for Higher-Resolution Adaptation},
  author={Guo, Lanqing and He, Yingqing and Chen, Haoxin and Xia, Menghan and Cun, Xiaodong and Wang, Yufei and Huang, Siyu and Zhang, Yong and Wang, Xintao and Chen, Qifeng and others},
  journal={arXiv preprint arXiv:2402.10491},
  year={2024}
}

📭 Contact

If your have any comments or questions, feel free to contact Lanqing Guo or Yingqing He.