README.md

metadata

license: other
license_name: license
license_link: LICENSE
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
dataset_info:
  features:
    - name: id
      dtype: string
    - name: date
      dtype: string
    - name: bbox
      sequence: float64
    - name: geometry
      struct:
        - name: coordinates
          sequence:
            sequence:
              sequence: float64
        - name: type
          dtype: string
    - name: satellite
      dtype: string
    - name: track
      dtype: string
    - name: direction
      dtype: string
    - name: mode
      dtype: string
    - name: band
      dtype: string
    - name: polarization
      sequence: string
    - name: azimuth_res
      dtype: float64
    - name: range_res
      dtype: float64
    - name: rows
      dtype: int64
    - name: cols
      dtype: int64
    - name: size
      dtype: int64
    - name: image_href
      dtype: string
    - name: image
      dtype:
        image:
          decode: false
    - name: __index_level_0__
      dtype: int64
  splits:
    - name: train
      num_bytes: 489980007
      num_examples: 1
  download_size: 473816647
  dataset_size: 489980007

AGPL-3.0 License: This OSI-approved open-source license is for research only activities. See the LICENSE file for more details.

Enterprise License: Designed for commercial use, this license permits integration of AI models derived from this dataset into commercial goods and services, bypassing the open-source requirements of AGPL-3.0. If your scenario involves embedding this dataset into a commercial offering, contact me.

Download json data (Last extraction 08/12/2024)

python extract.py

Download the GEC images and create an excel file with some metadata

python download.py

Download the label images from ESA WorldCover 2021

python label.py