SUIM dataset

Dataset for semantic segmentation of underwater imagery. 

 

suim_data
Object categorySymbolRGB color code
Background (waterbody)BW000 (black)
Human diversHD001 (blue)
Aquatic plants and sea-grassPF010 (green)
Wrecks and ruinsWR011 (sky)
Robots (AUVs/ROVs/instruments)RO100 (red)
Reefs and invertebratesRI101 (pink)
Fish and vertebratesFV110 (yellow)
Sea-floor and rocksSR111 (white)

Folder structure:

  • train_val/ contains 1525 paired samples for training/validation
    • images/: RGB images of underwater scenes
    • masks/: segmentation labels
      • Each RGB color represents a different object category
      • Further details in the paper (Section III) 
  • TEST/ contains 110 paired samples for benchmark evaluation
    • images/: RGB test images
    • masks/: ground truths labels
      • Combined RGB masks are provided
      • Individual binary masks are also provided in separate folders
  • Benchmark_Evaluation/ contains all the experimental data for the SOTA models