ryoungj/optdom

[ICLR'22] Self-supervised learning optimally robust representations for domain shift.

33
/ 100
Emerging

This project helps machine learning practitioners build more robust image classification models that perform well even when the real-world data they encounter is different from their training data. You input pre-trained image models (like CLIP) and image datasets from various 'domains' or sources. The output is a fine-tuned model that is better at generalizing to new, unseen data variations. This is for machine learning engineers and researchers who deploy models in diverse, changing environments.

No commits in the last 6 months.

Use this if your image classification models struggle with 'domain shift' – when data in deployment looks different from your training data, causing accuracy to drop.

Not ideal if you are looking for an out-of-the-box solution without needing to understand or adapt advanced machine learning techniques.

robust-ai image-classification domain-generalization computer-vision model-deployment
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 7 / 25
Maturity 16 / 25
Community 10 / 25

How are scores calculated?

Stars

25

Forks

3

Language

Python

License

MIT

Last pushed

Feb 02, 2022

Commits (30d)

0

Get this data via API

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

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