README.md

---
library_name: diffusers
license: mit
pipeline_tag: image-to-image
tags:
- computed-tomography
- ct-reconstruction
- diffusion-model
- inverse-problems
- dm4ct
- sparse-view-ct
---

# Pixel Diffusion UNet – LoDoInd (DM4CT)

This repository contains the pretrained **pixel-space diffusion UNet** used in the benchmark study **DM4CT: Benchmarking Diffusion Models for CT Reconstruction (ICLR 2026)**.

- **Paper:** [DM4CT: Benchmarking Diffusion Models for Computed Tomography Reconstruction](https://huggingface.co/papers/2602.18589)
- **ArXiv:** [https://arxiv.org/abs/2602.18589](https://arxiv.org/abs/2602.18589)
- **Project Page:** [https://dm4ct.github.io/DM4CT/](https://dm4ct.github.io/DM4CT/)
- **Codebase:** [https://github.com/DM4CT/DM4CT](https://github.com/DM4CT/DM4CT)

---

## 🔬 Model Overview

This model learns a **prior over CT reconstruction images** using a denoising diffusion probabilistic model (DDPM). It operates directly in **pixel space** (not latent space).

- **Architecture**: 2D UNet (Diffusers `UNet2DModel`)
- **Input resolution**: 512 × 512
- **Channels**: 1 (grayscale CT slice)
- **Training objective**: ε-prediction (standard DDPM formulation)
- **Noise schedule**: Linear beta schedule
- **Training dataset**: Industry CT dataset (LoDoInd)
- **Intensity normalization**: Rescaled to (-1, 1)

This model is intended to be combined with data-consistency correction for CT reconstruction tasks.

---

## 📊 Dataset: LoDoInd

Source: [LoDoInd on Zenodo](https://zenodo.org/records/10391412)

Preprocessing steps:
- Train/test split
- Rescale reconstructed slices to (-1, 1)
- No geometry information is embedded in the model

The model learns an unconditional image prior over CT slices.

---

## 🧠 Training Details

- **Optimizer:** AdamW
- **Learning rate:** 1e-4
- **Hardware:** NVIDIA A100 GPU
- **Training script:** [train_pixel.py](https://github.com/DM4CT/DM4CT/blob/main/train_pixel.py)

---

## 🚀 Usage

```python
from diffusers import DDPMPipeline

# Load the pipeline
pipeline = DDPMPipeline.from_pretrained("jiayangshi/lodoind_pixel_diffusion")
pipeline.to("cuda")

# Generate a CT slice prior
image = pipeline().images[0]
image.save("generated_ct_slice.png")
```

---

## Citation

```bibtex
@inproceedings{shi2026dmct,
title={{DM}4{CT}: Benchmarking Diffusion Models for Computed Tomography Reconstruction},
author={Shi, Jiayang and Pelt, Dani{\"{e}}l M and Batenburg, K Joost},
booktitle={The Fourteenth International Conference on Learning Representations},
year={2026},
url={https://openreview.net/forum?id=YE5scJekg5}
}
```