FutureXiang/ddae

[ICCV 2023 Oral] Official Implementation of "Denoising Diffusion Autoencoders are Unified Self-supervised Learners"

33
/ 100
Emerging

This project helps machine learning researchers to efficiently train and evaluate denoising diffusion autoencoders for both generating new images and extracting useful features from existing ones. Researchers can input a dataset of images and define a diffusion model architecture. The output includes trained models, generated image samples, and metrics like FID and classification accuracy for assessing model performance. This is for machine learning researchers and practitioners specializing in computer vision and generative models.

183 stars.

Use this if you are a machine learning researcher working with generative models and need to pre-train, sample from, and evaluate denoising diffusion autoencoders for image generation and feature learning.

Not ideal if you are looking for a plug-and-play tool for image generation or classification without needing to understand or configure advanced deep learning models and training processes.

deep-learning-research generative-ai computer-vision image-synthesis self-supervised-learning
No License No Package No Dependents
Maintenance 6 / 25
Adoption 10 / 25
Maturity 8 / 25
Community 9 / 25

How are scores calculated?

Stars

183

Forks

8

Language

Python

License

Last pushed

Dec 01, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/FutureXiang/ddae"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.