bahjat-kawar/time-diffusion
Official code repo for "Editing Implicit Assumptions in Text-to-Image Diffusion Models"
This tool helps researchers and engineers working with AI image generation models to correct unwanted biases or 'implicit assumptions' in how those models interpret text prompts. You input a text-to-image diffusion model and a dataset of prompts, and it outputs an edited model that generates images with modified or corrected characteristics based on your specified changes. This is for AI model developers and researchers who need to fine-tune generative AI behavior.
No commits in the last 6 months.
Use this if you need to precisely adjust how a text-to-image AI model interprets certain concepts or attributes, such as correcting gender stereotypes or cultural biases in generated images.
Not ideal if you are an end-user simply generating images and don't need to modify the underlying AI model's behavior.
Stars
87
Forks
2
Language
Python
License
—
Category
Last pushed
Mar 15, 2023
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/bahjat-kawar/time-diffusion"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
OPTML-Group/Unlearn-Saliency
[ICLR24 (Spotlight)] "SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in...
Shilin-LU/VINE
[ICLR 2025] "Robust Watermarking Using Generative Priors Against Image Editing: From...
WindVChen/DiffAttack
An unrestricted attack based on diffusion models that can achieve both good transferability and...
koninik/DiffusionPen
Official PyTorch Implementation of "DiffusionPen: Towards Controlling the Style of Handwritten...
Wuyxin/DISC
(ICML 2023) Discover and Cure: Concept-aware Mitigation of Spurious Correlation