---
license:
- odbl
- cc-by-sa-4.0
size_categories:
- n<1K
pretty_name: TerraMesh-Masks-Eval
tags:
- Earth observation
- Zero-Shot Segmentation
- Benchmark
task_categories:
- image-feature-extraction
library_name: webdataset
configs:
- config_name: default
  data_files:
  - split: test
    path: test_metadata.parquet
---

[![arXiv](https://img.shields.io/badge/arXiv-coming_soon-b31b1b?logo=arxiv)](https://arxiv.org/)
[![Code](https://img.shields.io/badge/GitHub-TerraMesh_Masks-0F62FE?logo=github)](https://github.com/IBM/TerraMesh-Masks)
[![HuggingFace](https://img.shields.io/badge/Hugging_Face-TerraMesh_Masks-FFD21E?logo=huggingface)](https://huggingface.co/ibm-esa-geospatial/TerraMesh-Masks)

# TerraMesh-Masks-Eval
 
TerraMesh-Masks-Eval is a human-verified benchmark dataset to evaluate open-vocabulary segmentation models on satellite imagery. This dataset provides binary segmentation masks with captions togehther with input samples from [TerraMesh](https://huggingface.co/ibm-esa-geospatial/TerraMesh).
We also provide a training dataset, called [TerraMesh-Masks](https://huggingface.co/ibm-esa-geospatial/TerraMesh-Masks).

Examples from the evaluation subset: 

![eval_examples.png](https://github.com/IBM/TerraMesh-Masks/raw/main/assets/eval_examples.png)


## Usage

Download the data loading code from [GitHub](https://github.com/IBM/TerraMesh-Masks) and install requirements with `pip install -r requirements.txt`. For development, you can stream data directly from Hugging Face.

Load from samples from the evaluation dataset:
```python
from torch.utils.data import DataLoader
from terramesh.masks_eval_dataset import build_eval_dataset

dataset = build_eval_dataset(
    path="https://huggingface.co/datasets/ibm-esa-geospatial/TerraMesh-Masks-Eval/resolve/main/test/", # Default
    modalities=["S2RGB", "BINARY_MASK"], # Select from ["S2L2A", "S2L1C", "S2RGB", "S1_merged", "DEM", "BINARY_MASK"]
    batch_size=4,
)
dataloader  = DataLoader(dataset, batch_size=None)
batch = next(iter(dataloader))
```

## Download

You can download the dataset with the Hugging Face CLI tool

```shell
hf download ibm-esa-geospatial/TerraMesh-Masks-Eval --repo-type dataset --local-dir data/TerraMesh
```

This results in the following data structure:
```text
TerraMesh
├── test
│   ├── BINARY_MASK.tar
│   ├── DEM.tar
│   ├── S1_merged.tar
│   ├── S2L1C.tar
│   ├── S2L2A.tar
│   └── S2RGB.tar
└── test_metadata.parquet
```

Different from TerraMesh, the samples are now stored as tif files which makes reviewing samples easier. You can also untar the archives and load them without using `webdataset`.

## Citation

Preprint for TerraMesh-Masks is coming soon!

TerraMesh:

```bibtex
@article{blumenstiel2025terramesh,
  title={Terramesh: A planetary mosaic of multimodal earth observation data},
  author={Blumenstiel, Benedikt and Fraccaro, Paolo and Marsocci, Valerio and Jakubik, Johannes and Maurogiovanni, Stefano and Czerkawski, Mikolaj and Sedona, Rocco and Cavallaro, Gabriele and Brunschwiler, Thomas and Bernabe-Moreno, Juan and others},
  journal={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
  year={2025},
}
```

## Licenses

The mask annotations are released under ODbL, following the OpenStreetMap license.

The image data is released under CC-BY-SA 4.0, following the license from TerraMesh.

## Acknowledgements

TerraMesh is part of the **FAST‑EO** project funded by the European Space Agency Φ‑Lab.

The binary annotations are based on the base layer provided by [Overture Maps](https://overturemaps.org/) and from [OpenStreetMap](https://www.openstreetmap.org/copyright) tags.

The image data in TerraMesh (and therefore this repository) is sourced from various sources:

The satellite images (S2L1C, S2L2A, S1GRD, S1RTC) are sourced from the [SSL4EO‑S12 v1.1](https://huggingface.co/datasets/embed2scale/SSL4EO-S12-v1.1) (CC-BY-4.0) and [MajorTOM‑Core](https://huggingface.co/Major-TOM) (CC-BY-SA-4.0) datasets.

The LULC data is provided by [ESRI, Impact Observatory, and Microsoft](https://planetarycomputer.microsoft.com/dataset/io-lulc-annual-v02) (CC-BY-4.0).

The DEM data is produced using [Copernicus WorldDEM-30](https://dataspace.copernicus.eu/explore-data/data-collections/copernicus-contributing-missions/collections-description/COP-DEM) © DLR e.V. 2010-2014 and © Airbus Defence and Space GmbH 2014-2018 provided under COPERNICUS by the European Union and ESA; all rights reserved