Residual Denoising Diffusion Models

This repository is the official implementation of Residual Denoising Diffusion Models.

Note:

The current setting is to train two unets (one to estimate the residuals and one to estimate the noise), which can be used to explore partially path-independent generation process.
Other tasks need to modify a) [self.alphas_cumsum[t]*self.num_timesteps, self.betas_cumsum[t]*self.num_timesteps]] -> [t,t] (in L852 and L1292). b) For image restoration, generation=False in L120, convert_to_ddim=False in L640. c) modify the corresponding experimental settings (see Table 4 in the Appendix).
The code is being updated.

Requirements

To install requirements:

conda env create -f install.yaml

Dataset

Raindrop

GoPro

ISTD

SID-RGB: kexu or download

LOL

CelebA

Training

To train RDDM, run this command:

python train.py

accelerate launch train.py

Evaluation

To evaluate image generation, run:

cd eval/image_generation_eval/
python fid_and_inception_score.py path_of_gen_img

For image restoration, MATLAB evaluation codes in ./eval.

Pre-trained Models

The pre-trained models will be provided later.

Results

See Table 3 in main paper.

Other experiments

We can convert a pre-trained DDIM to RDDM by coefficient transformation (see code).

Citation

If you find our work useful in your research, please consider citing:

@article{liu2023residual,
    title={Residual Denoising Diffusion Models}, 
    author={Jiawei Liu and Qiang Wang and Huijie Fan and Yinong Wang and Yandong Tang and Liangqiong Qu},
    year={2023},
    journal={arXiv preprint arxiv:2308.13712}
}

Contact

Please contact Jiawei Liu if there is any question (liujiawei18@mails.ucas.ac.cn).

XDeadlocked/RDDM