rvorias/ind_knn_ad

Vanilla torch and timm industrial knn-based anomaly detection for images.

48
/ 100
Emerging

This project helps quality control engineers, manufacturing specialists, and product inspectors automatically detect tiny defects or anomalies in product images. You input a collection of images of 'good' products, and it learns what's normal. Then, when you feed it new product images, it outputs an 'anomaly map' highlighting exactly where and how unusual an area in the image is, helping you quickly spot manufacturing flaws.

164 stars. No commits in the last 6 months.

Use this if you need to reliably identify subtle defects in visual inspections of manufactured goods, medical scans, or other image-based data, especially when anomalies are rare or hard for humans to spot.

Not ideal if your anomalies are obvious, if you have very little data for 'good' items, or if you need to classify multiple specific defect types rather than just flagging general anomalies.

quality-control manufacturing-inspection defect-detection visual-inspection product-screening
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 22 / 25

How are scores calculated?

Stars

164

Forks

50

Language

Python

License

MIT

Last pushed

Oct 04, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/rvorias/ind_knn_ad"

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