Dataset Card for HASYv2

Dataset Details

Dataset Description

This is a dataset of handwritten symbols similar to MNIST. It contains 168233 instances of 369 classes.

Dataset Sources

• Homepage: https://github.com/MartinThoma/HASY?tab=readme-ov-file
• Paper: Thoma, M. (2017). The hasyv2 dataset. arXiv preprint arXiv:1701.08380.

Dataset Structure

Total images: 168,233

Classes: 369 categories

Splits:

• Train: 151,241 images

• Test: 16,992 images

Image specs: PNG format, 32×32 pixels, Grayscale

Example Usage

Below is a quick example of how to load this dataset via the Hugging Face Datasets library.

from datasets import load_dataset 

# Load the dataset 
dataset = load_dataset("randall-lab/hasy-v2", split="train", trust_remote_code=True) 
# dataset = load_dataset("randall-lab/hasy-v2", split="test", trust_remote_code=True) 

# Access a sample from the dataset 
example = dataset[0] 
image = example["image"] 
label = example["label"] 

image.show() # Display the image 
print(f"Label: {label}")

Citation

BibTeX:

@article{thoma2017hasyv2, title={The hasyv2 dataset}, author={Thoma, Martin}, journal={arXiv preprint arXiv:1701.08380}, year={2017} }