Datasets:

isaaccorley
/

aid

text stringlengths 10 25
airport_10.jpg
airport_100.jpg
airport_101.jpg
airport_103.jpg
airport_106.jpg
airport_107.jpg
airport_108.jpg
airport_11.jpg
airport_110.jpg
airport_111.jpg
airport_114.jpg
airport_118.jpg
airport_119.jpg
airport_12.jpg
airport_120.jpg
airport_122.jpg
airport_123.jpg
airport_124.jpg
airport_126.jpg
airport_129.jpg
airport_13.jpg
airport_131.jpg
airport_133.jpg
airport_134.jpg
airport_135.jpg
airport_139.jpg
airport_143.jpg
airport_146.jpg
airport_147.jpg
airport_148.jpg
airport_150.jpg
airport_151.jpg
airport_153.jpg
airport_154.jpg
airport_157.jpg
airport_158.jpg
airport_16.jpg
airport_160.jpg
airport_162.jpg
airport_165.jpg
airport_166.jpg
airport_169.jpg
airport_17.jpg
airport_170.jpg
airport_171.jpg
airport_172.jpg
airport_173.jpg
airport_174.jpg
airport_175.jpg
airport_177.jpg
airport_178.jpg
airport_179.jpg
airport_18.jpg
airport_180.jpg
airport_181.jpg
airport_182.jpg
airport_183.jpg
airport_188.jpg
airport_189.jpg
airport_19.jpg
airport_190.jpg
airport_191.jpg
airport_192.jpg
airport_194.jpg
airport_195.jpg
airport_196.jpg
airport_197.jpg
airport_20.jpg
airport_201.jpg
airport_202.jpg
airport_203.jpg
airport_204.jpg
airport_205.jpg
airport_206.jpg
airport_208.jpg
airport_211.jpg
airport_212.jpg
airport_213.jpg
airport_215.jpg
airport_216.jpg
airport_217.jpg
airport_22.jpg
airport_220.jpg
airport_222.jpg
airport_223.jpg
airport_224.jpg
airport_227.jpg
airport_228.jpg
airport_23.jpg
airport_235.jpg
airport_236.jpg
airport_237.jpg
airport_24.jpg
airport_240.jpg
airport_242.jpg
airport_243.jpg
airport_247.jpg
airport_248.jpg
airport_25.jpg
airport_250.jpg

End of preview. Expand in Data Studio

Redistributed without modification from https://huggingface.co/datasets/blanchon/AID, itself a rehost of Xia et al. 2017's Aerial Image Dataset. 10,000 RGB images, 600x600px, 30 scene classes, extracted from Google Earth.

License is unspecified upstream -- neither the original paper nor the blanchon/AID rehost states one. Verify suitability for your use case before redistributing further.

Please cite https://doi.org/10.1109/TGRS.2017.2685945 if you use this dataset.

The train/val/test split files (aid-{train,val,test}.txt) are not official -- no published AID split exists. They are a deterministic, per-class stratified 60/20/20 split (seed 0 over sorted filenames), generated by torchgeo-bench's scripts/generate_aid_splits.py.

Downloads last month: 73