wogong/pytorch-dann

A PyTorch implementation for Unsupervised Domain Adaptation by Backpropagation

41
/ 100
Emerging

This project helps machine learning engineers or researchers adapt a trained image classification model from one domain to another without needing new labels for the target domain. You input a pre-trained model on a 'source' image dataset and an unlabeled 'target' image dataset, and it outputs a refined model that performs better on the target domain. This is for professionals working with computer vision tasks where collecting labeled data for every new scenario is impractical or too expensive.

151 stars. No commits in the last 6 months.

Use this if you have a well-performing image classification model on one dataset but need to apply it to a similar dataset where it currently struggles due to visual differences, and you don't have labeled data for the new set.

Not ideal if you have a completely new classification problem or if you have ample labeled data for your target domain, as a direct training approach might be more suitable.

computer-vision image-classification machine-learning-engineering model-adaptation unsupervised-learning
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 15 / 25

How are scores calculated?

Stars

151

Forks

18

Language

Jupyter Notebook

License

MIT

Last pushed

Sep 30, 2020

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/wogong/pytorch-dann"

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