amazon-science/crossnorm-selfnorm

CrossNorm and SelfNorm for Generalization under Distribution Shifts, ICCV 2021

34
/ 100
Emerging

This project helps machine learning engineers and researchers improve the reliability of image classification models. When your model encounters new images that are slightly different from its training data, it can struggle to maintain accuracy. This provides techniques that take your existing image classification models and training data to produce more robust models that perform better on corrupted or shifted data. This is for machine learning practitioners building and deploying computer vision models.

128 stars. No commits in the last 6 months.

Use this if you need to build image classification systems that are resilient to common data corruptions or shifts, such as noise, blur, or digital distortions in the input images.

Not ideal if you are looking for a pre-trained, ready-to-use image classification API or a tool for general machine learning tasks beyond deep learning model robustness.

computer-vision image-classification deep-learning-robustness model-generalization pytorch
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 8 / 25

How are scores calculated?

Stars

128

Forks

6

Language

Python

License

Apache-2.0

Last pushed

Sep 10, 2021

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/amazon-science/crossnorm-selfnorm"

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