snap-research/InstantRestore
Official Implementation for "InstantRestore: Single-Step Personalized Face Restoration with Shared-Image Attention"
This tool quickly and accurately restores degraded facial images, making them look clear while keeping the person's unique features intact. You provide a blurry or damaged face photo along with a few clear reference photos of the same person, and it outputs a high-quality, restored image almost instantly. It's ideal for anyone needing to clean up old photos or enhance low-quality facial captures.
133 stars. No commits in the last 6 months.
Use this if you need to restore severely degraded face images quickly and ensure the restored image perfectly preserves the original person's identity and facial details.
Not ideal if you need to restore non-face images or if you don't have any clear reference images of the person whose face you want to restore.
Stars
133
Forks
5
Language
Python
License
—
Category
Last pushed
Sep 23, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/snap-research/InstantRestore"
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