Darth-Kronos/Unsupervised-Domain-Adaptation

Empirical evaluation and analysis of state-of-the-art methods for unsupervised domain adaptation on OFFICE-31 dataset, a benchmark dataset for visual domain adaptation.

31
/ 100
Emerging

This project offers a way for machine learning engineers to adapt their image classification models to new, unlabeled datasets. It takes an existing model trained on one set of images (like product photos from Amazon) and fine-tunes it to perform well on a different but related set of images (like photos from a webcam or DSLR), even without new labels. This helps maintain model performance when the real-world data distribution shifts.

No commits in the last 6 months.

Use this if you have an image classification model that performs poorly when deployed in a new environment or with a new type of visual data because the data distribution is different from your original training data, and you lack new labeled data.

Not ideal if your problem does not involve visual data, or if you have plenty of labeled data available for your target domain.

computer-vision image-classification model-adaptation machine-learning-deployment data-drift
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 10 / 25

How are scores calculated?

Stars

14

Forks

2

Language

Jupyter Notebook

License

MIT

Last pushed

Jul 21, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/Darth-Kronos/Unsupervised-Domain-Adaptation"

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