Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Official Pytorch Implementation of the paper: Wavelet Diffusion Models are fast and scalable Image Generators (CVPR'23)
| Date | Stars |
|---|---|
| 2026-07-31 | 442 |
| 2026-08-01 | 442 |
| 2026-08-06 | 442 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
##### Table of contents
1. [Installation](#installation)
2. [Dataset preparation](#dataset-preparation)
3. [How to run](#how-to-run)
4. [Results](#results)
5. [Evaluation](#evaluation)
6. [Acknowledgments](#acknowledgments)
7. [Contacts](#contacts)
# Official PyTorch implementation of "Wavelet Diffusion Models are fast and scalable Image Generators" [(CVPR'23)](https://arxiv.org/abs/2211.16152)
<div align="center">
<a href="https://hao-pt.github.io/" target="_blank">Hao Phung</a>   <b>·</b>  
<a href="https://scholar.google.com/citations?user=g0RS3_kAAAAJ&hl=en&oi=ao" target="_blank">Quan Dao</a>   <b>·</b>  
<a href="https://sites.google.com/site/anhttranusc/" target="_blank">Anh Tran</a>
<br> <br>
<a href="https://www.vinai.io/">VinAI Research</a>
<br> <br>
<a href="https://arxiv.org/abs/2211.16152">[Paper]</a>   
<a href="https://drive.google.com/file/d/1LSEYfdhS4Zjtx1VRrctmVt6xjEjgmpVA/view?usp=sharing">[Poster]</a>   
<a href="https://drive.google.com/file/d/11JE-RFtYJWx6XdXH8zZxgzRAvGJ6-IV2/view?usp=sharing">[Slides]</a>   
<a href="https://youtu.be/KaIMMamhKsU">[Video]</a>
</div>
<br>
<div align="center">
<img width="1000" alt="teaser" src="assets/single_wavelet.png"/>
</div>
WaveDiff is a novel wavelet-based diffusion scheme that employs low-and-high frequency components of wavelet subbands from both image and feature levels. These are adaptively implemented to accelerate the sampling process while maintaining good generation quality. Experimental results on CelebA-HQ, CIFAR-10, LSUN-Church, and STL-10 datasets show that WaveDiff provides state-of-the-art training and inference speed, which serves as a stepping-stone to offering real-time and high-fidelity diffusion model.
Details of the model architecture and experimental results can be found in [our following paper](https://arxiv.org/abs/2211.16152):
```bibtex
@InProceedings{phung2023wavediff,
author = {Phung, Hao and Dao, Quan and Tran, Anh},
title = {Wavelet Diffusion Models Are Fast and Scalable Image Generators},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2023},
pages = {10199-10208}
}
```
**Please CITE** our paper whenever this repository is used to help produce published results or incorporated into other software.
## Installation ##
Python `3.7.13` and Pytorch `1.10.0` are used in this implementation.
It is recommended to create `conda` env from our provided [environment.yml](./environment.yml):
```
conda env create -f environment.yml
conda activate wavediff
```
Or you can install neccessary libraries as follows:
```bash
pip install -r requirements.txt
```
For `pytorch_wavelets`, please follow [here](https://github.com/fbcotter/pytorch_wavelets.git).
## Dataset preparation ##
We trained on four datasets, including CIFAR10, STL10, LSUN Church Outdoor 256 and CelebA HQ (256 & 512).
For CIFAR10 and STL10, they will be automatically downloaded in the first time execution.
For CelebA HQ (256) and LSUN, please check out [here](https://github.com/NVlabs/NVAE#set-up-file-paths-and-data) for dataset preparation.
For CelebA HQ (512 & 1024), please download two zip files: [data512x512.zip](https://drive.google.com/file/d/1E23HCNL-v9c54Wnzkm9yippBW8IaLUXp/view?usp=share_link) and [data1024x1024.zip](https://drive.google.com/file/d/1-LFFkFKNuyBO1sjkM4t_AArIXr3JAOyl/view?usp=share_link) and then generate [LMDB](https://lmdb.readthedocs.io/en/release/) format dataset by [Torch Toolbox](https://github.com/PistonY/torch-toolbox#5-make-and-use-lmdb-dataset).
> Those two links of high-res data seem to be broken so we provide our processed lmdb files at [here](https://drive.google.com/drive/folders/1d_8dhc3HFsSYTvQNN7vbDv5D_lRGHm4g?usp=sharing).
Once a dataset is downloaded, please put it in `data/` directory as follows:
```
data/
├── SExcerpt of 7,793 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:18feab713f200134, topic:diffusion-models