Mars-Bench Surface Change Detection, MRO CTX Before/After Pairs
- Provider
- Mirali Purohit, Hannah Kerner et al. (Mars-Bench, Arizona State University)
- License
- CC-BY-4.0 (source)
- Size
- 11 files
- Format
- CSV, Parquet (in sampled files)
- Stored at
- Hugging Face (https://huggingface.co/datasets/Mirali33/mb-change_cls_ctx)
- Access
- Free and public. No account with the data provider is needed.
- DOI
- 10.48550/arXiv.2510.24010
- Category
- Astronomy & Space
- Source
- https://huggingface.co/datasets/Mirali33/mb-change_cls_ctx
About Mars-Bench Surface Change Detection, MRO CTX Before/After Pairs
Binary change detection on Mars from the MRO Context Camera (CTX): each example is a 150 x 150 pixel composite of a before and an after image of the same ground, following Kerner et al. (2019), with the before image in the blue channel, the after image in green and red set to zero. Labels are change or no_change. The original release had no standard split, so Mars-Bench generated one (36 training, 10 validation and 10 test pairs in the paper's table). It is a very small, hard benchmark meant for testing transfer and few-shot behavior of pretrained vision models. Parquet files under data/ hold the composite image and label, with annotation.csv and mapping.json; load with datasets. Under 2 MB. Part of Mars-Bench (Purohit et al., NeurIPS 2025 Datasets and Benchmarks track, arXiv:2510.24010), 20 standardized Mars datasets released by the authors under CC BY 4.0; cite Mars-Bench and Kerner et al. (2019).
How to load Mars-Bench Surface Change Detection, MRO CTX Before/After Pairs in Python
import pandas as pd
df = pd.read_csv("hf://datasets/Mirali33/mb-change_cls_ctx/annotation.csv")
df.head()Data structure
File types in sampled files: CSV, Parquet
Files
Example file: annotation.csv
Analyze this dataset without downloading it
On Scigantic, Mars-Bench Surface Change Detection, MRO CTX Before/After Pairs opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.
Related Astronomy & Space datasets
- TIMED Mission Upper Atmosphere Data: SABER, GUVI, SEE and TIDI (CDAWeb on HelioCloud)
- Sup3rCC
- Multimodal Universe: SDSS DR17 Optical Spectra
- Mars-Bench Atmospheric Dust Classification, HiRISE RDR Tiles (Dusty vs Not Dusty)
- ExoMiner++ TESS 2-Minute Threshold-Crossing Events (TCEs), Sectors 1-67: Vetting Features and Labels for 204,729 Transit Signals
- Multimodal Universe: TESS SPOC Light Curves for 1.1 Million Stars (HATS)
- Multimodal Universe: Galaxy10 DECaLS Galaxy Morphology Images (17,736 Galaxies, HATS)
- Lunar South Pole Synthetic Orthographic Renders for Terrain Relative Navigation (LOLA DEMs, 60-Sun Grid)