MiZhenxing/ThinkDiff
ICML2025, I Think, Therefore I Diffuse: Enabling Multimodal In-Context Reasoning in Diffusion Models
This project helps generate images from complex, multi-part descriptions that involve reasoning, rather than just simple text prompts. You provide a mix of images and text, and it produces a new image that reflects the logical connections and visual elements you specified. This is useful for designers, marketers, and content creators who need to generate visuals based on nuanced instructions.
192 stars. No commits in the last 6 months.
Use this if you need to create images that demonstrate sophisticated understanding of multiple inputs, such as generating a video from a single image and text, or combining elements from several images and text descriptions into one new image.
Not ideal if you are looking for a simple text-to-image generator without complex reasoning or multimodal inputs.
Stars
192
Forks
8
Language
Python
License
MIT
Category
Last pushed
Sep 07, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/MiZhenxing/ThinkDiff"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
FlorianFuerrutter/genQC
Generative Quantum Circuits
horseee/DeepCache
[CVPR 2024] DeepCache: Accelerating Diffusion Models for Free
Gen-Verse/MMaDA
MMaDA - Open-Sourced Multimodal Large Diffusion Language Models (dLLMs with block diffusion,...
kuleshov-group/mdlm
[NeurIPS 2024] Simple and Effective Masked Diffusion Language Model
Shark-NLP/DiffuSeq
[ICLR'23] DiffuSeq: Sequence to Sequence Text Generation with Diffusion Models