index/label_map.json

{
  "classification_classes": [
    "bg",
    "diabetic",
    "no_wound",
    "pressure",
    "surgical",
    "toes",
    "venous"
  ],
  "label_to_id": {
    "bg": 0,
    "diabetic": 1,
    "no_wound": 2,
    "pressure": 3,
    "surgical": 4,
    "toes": 5,
    "venous": 6
  },
  "id_to_label": {
    "0": "bg",
    "1": "diabetic",
    "2": "no_wound",
    "3": "pressure",
    "4": "surgical",
    "5": "toes",
    "6": "venous"
  },
  "notes": [
    "Union label space across datasets. Missing classes in a dataset are handled during training via label masking."
  ]
}