Shilin-LU/MACE
[CVPR 2024] "MACE: Mass Concept Erasure in Diffusion Models" (Official Implementation)
This project helps anyone working with AI image generation models to control the content they produce. It allows you to remove specific "concepts"—like objects, celebrities, explicit content, or artistic styles—from a diffusion model's ability to generate images. You input a list of unwanted concepts, and the system modifies the model so it no longer generates images embodying those concepts, even when prompted, while preserving the ability to create unrelated images. This is for AI artists, content moderators, or anyone managing image generation to ensure safe and appropriate outputs.
393 stars. No commits in the last 6 months.
Use this if you need to prevent large-scale text-to-image AI models from generating images of specific unwanted concepts, such as copyrighted material, explicit content, or particular artistic styles.
Not ideal if you only need to remove a single, simple concept or if you are looking for a tool to enhance or add new concepts to an image generation model.
Stars
393
Forks
31
Language
Jupyter Notebook
License
—
Category
Last pushed
Jun 02, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/Shilin-LU/MACE"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
huggingface/diffusers
🤗 Diffusers: State-of-the-art diffusion models for image, video, and audio generation in PyTorch.
bghira/SimpleTuner
A general fine-tuning kit geared toward image/video/audio diffusion models.
mcmonkeyprojects/SwarmUI
SwarmUI (formerly StableSwarmUI), A Modular Stable Diffusion Web-User-Interface, with an...
nateraw/stable-diffusion-videos
Create 🔥 videos with Stable Diffusion by exploring the latent space and morphing between text prompts
TheDesignFounder/DreamLayer
Benchmark diffusion models faster. Automate evals, seeds, and metrics for reproducible results.