kensanata/numbers

Handwritten digits, a bit like the MNIST dataset.

45
/ 100
Emerging

This project provides two distinct datasets of handwritten and machine-printed digits, totaling over 800,000 images, including a unique collection of Swiss handwriting. It offers raw image files of digits, some with demographic metadata, for training and testing machine learning models. Researchers, data scientists, and educators working on optical character recognition or image classification tasks would find this useful.

No commits in the last 6 months.

Use this if you need a large, diverse dataset of digit images, especially if you are interested in regional handwriting variations or working with real-world document data that might include both handwritten and printed numbers.

Not ideal if you require a perfectly clean, uniformly distributed dataset exclusively of high-quality handwritten digits with precise author information, as some sets include printed digits, miscategorizations, or lack metadata.

handwriting-recognition optical-character-recognition image-classification machine-learning-datasets computer-vision
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 9 / 25
Maturity 16 / 25
Community 20 / 25

How are scores calculated?

Stars

95

Forks

25

Language

License

Last pushed

Jun 27, 2020

Commits (30d)

0

Get this data via API

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

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