IllustrisTNG SKIRT SDSS: Synthetic Galaxy Images for Morphology Tasks

Preprocessed synthetic galaxy images derived from the IllustrisTNG cosmological simulations. Raw multi-band FITS images were produced with the SKIRT Monte Carlo radiative transfer code in SDSS photometric bands and processed into 128 × 128 RGB images. The dataset was created by HITS-AIN and is designed as training data for the Spherinator/HiPSter framework.

Use Cases

• Train galaxy morphology classifiers based on synthetic RGB images.
• Develop generative models for galaxies using preprocessed simulation data.
• Benchmark computer vision algorithms on standardized 128x128 astronomical images.
• Study the relationship between simulated physical properties and visual morphology.

Strengths

• Images are preprocessed into a consistent 128x128 RGB format ready for machine learning.
• Data is derived from the established IllustrisTNG cosmological simulation suite.
• Images are produced using the specialized SKIRT radiative transfer code for physical accuracy.

Limitations

• The exact number of images (row count) is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect the specific assumptions and biases inherent to the IllustrisTNG simulation.

Provenance

Source

HITS-AIN

Collection Method

Derived from IllustrisTNG cosmological simulations, with images generated using the SKIRT Monte Carlo radiative transfer code.

Freshness

Last updated 2026-05-07 14:00:32; freshness should be verified.