hendrycks/robustness

Corruption and Perturbation Robustness (ICLR 2019)

48
/ 100
Emerging

This project helps machine learning engineers and researchers evaluate how robust their image classification models are when faced with common real-world image distortions like blur, noise, or adverse weather conditions. You provide your trained neural network model, and it outputs detailed metrics showing how well the model performs under various corruptions and perturbations. This is ideal for anyone developing or deploying computer vision systems, especially in critical applications where image quality might be unpredictable.

1,139 stars. No commits in the last 6 months.

Use this if you need to rigorously test your image classification model's reliability against common real-world image corruptions and perturbations.

Not ideal if you are looking for a tool to train or fine-tune your model, as this project focuses specifically on evaluation.

image-classification model-evaluation computer-vision machine-learning-engineering model-robustness
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 22 / 25

How are scores calculated?

Stars

1,139

Forks

151

Language

Python

License

Apache-2.0

Last pushed

Aug 24, 2022

Commits (30d)

0

Get this data via API

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

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