Instructions to use R-J/StainFuser with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use R-J/StainFuser with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("R-J/StainFuser", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
metadata
license: apache-2.0
datasets:
- R-J/SPI-2M
library_name: diffusers
tags:
- medical
StainFuser: Controlling Diffusion for Faster Neural Style Transfer in Multi-Gigapixel Histology Images
Repo containing model weights from the paper, Github code can be found here
Organisation
- checkpoint: StainFuser trained weights trained at 512x512 resolution with mixed magnification
- training: contains SD per-trained weights for backbone initialistaion in training
Citation
@misc{jewsbury2024stainfuser,
title={StainFuser: Controlling Diffusion for Faster Neural Style Transfer in Multi-Gigapixel Histology Images},
author={Robert Jewsbury and Ruoyu Wang and Abhir Bhalerao and Nasir Rajpoot and Quoc Dang Vu},
year={2024},
eprint={2403.09302},
archivePrefix={arXiv},
primaryClass={eess.IV}
}
