FAIR1M Dataset Overview

Basic Information

• Language: English
• License: Unknown
• Task Types:
  • Object Detection
• Tags:
  • Remote Sensing
  • Earth Observation
  • Geospatial
  • Satellite Imagery
  • Object Detection
• Dataset ID: FAIR1M
• Readable Name: FAIR1M

Dataset Description

• Overview: FAIR1M is a fine‑grained object recognition and detection dataset focusing on high‑resolution (0.3‑0.8 m) RGB images captured by Gaofen satellites and extracted from Google Earth. It contains 15 000 high‑resolution images covering a variety of objects and scenes. Annotations are provided as rotated bounding boxes for objects belonging to five major categories (ships, vehicles, aircraft, ball‑fields, roads) and 37 sub‑categories.
• Object Instance Count: 1 million
• Sample Count: 15 000
• Bands: 3 (RGB)
• Image Size: 1024×1024
• Image Resolution: 0.3–0.8 m
• Object Classes: 37
• Categories: 5 major categories, 37 sub‑categories
• Scene Categories: Passenger ship, motorboat, fishing vessel, tugboat, other vessels, construction ship, oil tanker, dry cargo ship, warship, small car, bus, truck, dump truck, other vehicles, van, trailer, tractor, excavator, tractor‑trailer, Boeing 737, Boeing 747, Boeing 777, Boeing 787, ARJ21, C919, A220, A321, A330, A350, other aircraft, baseball field, basketball court, football field, tennis court, roundabout, intersection, bridge
• Source: Gaofen / Google Earth

Usage

• Load Dataset: Use datasets.load_dataset("blanchon/FAIR1M") to load the dataset.

Citation

• APA: Sun, X., Wang, P., Yan, Z., Xu, F., Wang, R., Diao, W., Chen, J., Li, J., Feng, Y., Xu, T., Weinmann, M., Hinz, S., Wang, C., & Fu, K. (2021). FAIR1M: A Benchmark Dataset for Fine‑grained Object Recognition in High‑Resolution Remote Sensing Imagery. ISPRS Journal of Photogrammetry and Remote Sensing. https://doi.org/10.1016/j.isprsjprs.2021.12.004
• BibTeX:

  @article{sun2021fair1m,
      title     = {FAIR1M: A Benchmark Dataset for Fine‑grained Object Recognition in High‑Resolution Remote Sensing Imagery},
      author    = {Xian Sun and Peijin Wang and Zhiyuan Yan and F. Xu and Ruiping Wang and W. Diao and Jin Chen and Jihao Li and Yingchao Feng and Tao Xu and M. Weinmann and S. Hinz and Cheng Wang and K. Fu},
      journal   = {ISPRS Journal of Photogrammetry and Remote Sensing},
      year      = {2021},
      doi       = {10.1016/j.isprsjprs.2021.12.004},
      bibSource = {Semantic Scholar https://www.semanticscholar.org/paper/6d3c2dc63ff0deec10f60e5a515c93af4f8676f2}
  }