muzishen/IMAGHarmony

🧩 IMAGHarmony 🧩: Controllable image editing with consistent object quantity and layout. A structure-aware framework that ensures high fidelity and coherence in complex multi-object edits. It integrates harmony-aware attention and preference-guided noise selection to enable precise, stable, and semantically aligned generation.

47
/ 100
Emerging

Builds on Stable Diffusion XL and IP-Adapter, employing a harmony-aware module to jointly encode object structure and semantic information alongside preference-guided noise initialization for stabilized generation. The framework is trained and evaluated on HarmonyBench, a curated benchmark dataset with diverse multi-object editing scenarios, supporting dual-category edits and fine-grained control over object count and spatial arrangement. Provides both training and inference code with a Gradio demo interface for interactive testing.

607 stars.

No Package No Dependents
Maintenance 6 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 15 / 25

How are scores calculated?

Stars

607

Forks

45

Language

Python

License

Apache-2.0

Last pushed

Oct 18, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/muzishen/IMAGHarmony"

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