covisionlab/diffusion_labeling

Official implementation of "Bounding Box-Guided Diffusion for Synthesizing Industrial Images and Segmentation Map" accepted at Synthetic Data for Computer Vision Workshop - CVPR 2025

22
/ 100
Experimental

This project helps quality control engineers and manufacturing professionals create synthetic industrial images and their corresponding defect maps. It takes simple bounding box annotations (e.g., specifying where a scratch or dent should be) and generates realistic images of industrial products, along with precise segmentation masks highlighting those 'defects'. This allows for the creation of vast, varied datasets needed to train automated defect detection systems without needing real faulty products.

No commits in the last 6 months.

Use this if you need to generate high-quality synthetic images of industrial products with specific, localized defects and their precise segmentation maps, especially for training computer vision models.

Not ideal if you're looking for a simple, out-of-the-box solution that doesn't require technical setup or if your primary need is general image generation without specific defect localization guidance.

industrial-quality-control defect-detection synthetic-data-generation manufacturing-inspection computer-vision-training
No License Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 5 / 25
Maturity 8 / 25
Community 7 / 25

How are scores calculated?

Stars

10

Forks

1

Language

Python

License

Last pushed

Aug 28, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/covisionlab/diffusion_labeling"

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