CeMOS-IS/GenFormer

[ICPR 2024] Official repository of the paper "GenFormer - Generated Images are All You Need to Improve Robustness of Transformers on Small Datasets"

27
/ 100
Experimental

This project helps machine learning practitioners improve the reliability of image classification models, especially when working with limited real-world image datasets. By generating synthetic images and integrating them into the training process, it makes 'vision transformer' models more accurate and robust against unexpected variations. The output is a more dependable image classification model that can generalize better to new, unseen images, ideal for machine learning engineers or researchers building computer vision systems.

No commits in the last 6 months.

Use this if you need to train robust image classification models with limited real-world image data and want to leverage synthetic data generation to enhance performance and resilience.

Not ideal if you are working with non-image data or if your existing image datasets are already very large and diverse, as the benefits might be less pronounced.

image-classification computer-vision machine-learning-engineering synthetic-data model-robustness
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 6 / 25

How are scores calculated?

Stars

14

Forks

1

Language

Python

License

Apache-2.0

Last pushed

Aug 30, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/CeMOS-IS/GenFormer"

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