metadata
license: cc0-1.0
Tessera Sample Dataset
This dataset is a small sample of the Tessera dataset developed by the University of Cambridge Earth Observation Group.
This sample is intended for testing workflows, experimentation, and demonstration purposes.
Original Tessera dataset: GitHub Repository
Dataset Overview
TESSERA (Temporal Embeddings of Surface Spectra for Earth Representation and Analysis) encodes spatio-temporal information from Earth observation data into compact embeddings.
This sample dataset provides:
- Global coverage at 0.1° resolution
- Time-specific embeddings per grid tile (year 2024)
- 128-dimensional vector per pixel, capturing spectral, spatial, and temporal properties
- Grid-based organization, allowing easy selection of spatial subsets
Dataset Structure
The dataset is organized as follows:
tessera_sample/
├── train/
│ └── global_0.1_degree_representation/
│ └── 2024/
│ ├── grid_-94.65_40.75/
│ │ └── grid_-94.65_40.75_2024.tiff
│ ├── grid_-94.75_40.75/
│ │ └── grid_-94.75_40.75_2024.tiff
│ └── grid_-94.85_40.75/
│ └── grid_-94.85_40.75_2024.tiff
└── val/
└── global_0.1_degree_representation/
└── 2024/
└── grid_-94.95_40.75/
└── grid_-94.95_40.75_2024.tiff
- Each folder corresponds to a geospatial grid tile.
- File naming format:
grid_<longitude>_<latitude>_<year>.tiff.