mrJordi0/galaxy-zoo-dataset

Galaxy Morphological Classification (Derived from Galaxy Zoo)

Dataset Summary

This dataset is designed for Galaxy Morphological Classification tasks, specifically tailored for training deep learning models such as Efficient Vision Transformers (ViT).

The dataset contains galaxy images and their corresponding morphological labels (8 classes). The underlying images and ground-truth classifications are derived from the Galaxy Zoo DECaLS project (Walmsley et al., 2022). The 8-class morphological taxonomy and classification approach used in this dataset are based on Lin et al. (2021), which first applied Vision Transformers to galaxy morphology classification using this same data. This repository formats the original scientific data into a clean structure ready for Machine Learning pipelines.

• Original Data Source: Galaxy Zoo Data Releases
• Data Paper: Walmsley et al. (2022) — Galaxy Zoo DECaLS: Detailed Visual Morphology for 314,000 Galaxies
• Classification Approach Paper: Lin et al. (2021) — Galaxy Morphological Classification with Efficient Vision Transformer

Dataset Structure

The dataset consists of galaxy images (cutouts) and their corresponding integer class labels.

Data Fields

• image: A PIL.Image.Image object containing the galaxy cutout.
• label: An int64 integer representing the morphological class.

Label Mapping (Class ID to Morphology)

ID	Class Name	Description
0	Round Elliptical	Smooth galaxy, completely round shape.
1	In-between Elliptical	Smooth galaxy, slightly elongated.
2	Cigar-shaped Elliptical	Smooth galaxy, very elongated/cigar-like.
3	Edge-on Spiral	Spiral galaxy viewed from the side (disk visible).
4	Barred Spiral	Spiral galaxy with a central bar structure.
5	Unbarred Spiral	Spiral galaxy without a central bar.
6	Irregular	Galaxy with no defined shape or disturbed structure.
7	Merger	Two or more galaxies interacting/merging.

Visual Examples of the Classes:

Usage

from datasets import load_dataset

dataset = load_dataset("mrJordi0/galaxy-zoo-dataset")

train_data = dataset['train']
test_data = dataset['test']

id2label = {
    0: "Round Elliptical",
    1: "In-between Elliptical",
    2: "Cigar-shaped Elliptical",
    3: "Edge-on Spiral",
    4: "Barred Spiral",
    5: "Unbarred Spiral",
    6: "Irregular",
    7: "Merger"
}

example = train_data[0]
print(f"Label ID: {example['label']} | Morphology: {id2label[example['label']]}")
example['image'].show()

Citation

If you use this dataset, please cite both papers:

@misc{Lin2021,
    author = {Lin, Joshua Yao-Yu and Liao, Song-Mao and Huang, Hung-Jin and Kuo, Wei-Ting and Ou, Olivia Hsuan-Min},
    title = {Galaxy Morphological Classification with Efficient Vision Transformer},
    year = {2021},
    eprint = {2110.01024},
    archivePrefix = {arXiv},
    primaryClass = {astro-ph.GA},
    url = {https://arxiv.org/abs/2110.01024}
}

@article{Walmsley_2022,
    author = {Walmsley, Mike and Lintott, Chris and Géron, Tobias and Kruk, Sandor and Krawczyk, Coleman and Willett, Kyle W. and Bamford, Steven and Dickinson, Hugh and Fortson, Lucy and Gal, Yarin and Keel, William and Masters, Karen L. and Mehta, Vihang and Simmons, Brooke D. and Smethurst, Rebecca and Baeten, Elisabeth M L and Macmillan, Christine},
    title = {Galaxy Zoo DECaLS: Detailed visual morphology measurements from volunteers and deep learning for 314,000 galaxies},
    journal = {Monthly Notices of the Royal Astronomical Society},
    volume = {509},
    number = {3},
    pages = {3966-3988},
    year = {2022},
    doi = {10.1093/mnras/stab2093},
    url = {https://doi.org/10.1093/mnras/stab2093}
}

Acknowledgments

This dataset is a processed version created for machine learning applications. All credit for the original classifications and data collection goes to the Galaxy Zoo team and the hundreds of thousands of volunteers who participated in the project.

• Data Source: Galaxy Zoo Data Releases
• License: Creative Commons Attribution 4.0 International (CC BY 4.0)