KeyError: 'thing_dataset_id_to_contiguous_id'

[david95j2]

Hi guys, i am trying to register and train a custom panoptic dataset. I have converted the annotations to coco format.
I have only one concretcrack class and then the following is the format of the panoptic annotations

{
 "info": {...},
 "images": [
    {
      "id": 263,
      "width": 4032,
      "height": 3024,
      "file_name": "000263.jpg"
    }, ...
    ... 
  ],
  "annotations": [
    {
      "image_id": 263,
      "file_name": "000263_GT.png",
      "segments_info": [
        {
          "id": 39372,
          "category_id": 1,
          "bbox": [
            0.0,
            803.0,
            4032.0,
            2221.0
          ],
          "iscrowd": 0,
          "area" : 128846
        }
      ]
    }, ...
    ...
    "categories": [
       {
          "supercateory": "korad",
          "isthing" : 1,
          "id" : 1,
          "name" : "ConcreteCrack",
          "color": [
               0,
              153,
              204
          ]
       }
     ]
}

Instructions To Reproduce the Issue:

from detectron2.data.datasets import register_coco_panoptic, register_coco_panoptic_separated, load_coco_json
from detectron2.data import DatasetCatalog, MetadataCatalog
from detectron2.data.datasets.coco_panoptic import load_coco_panoptic_json

if 'ConcreteCrack' in DatasetCatalog.list():
    DatasetCatalog.remove('ConcreteCrack')

register_coco_panoptic("ConcreteCrack",{},"./datasets/crack/train2022/","./datasets/crack/panoptic_train2022/",
"./datasets/crack/annotations/panoptic_train2022.json","./datasets/crack/annotations/instances_train2022.json")

MetadataCatalog.get("ConcreteCrack").thing_classes = ["ConcreteCrack"]

dicts = load_coco_panoptic_json('./datasets/crack/annotations/panoptic_train2022.json','./datasets/crack/train2022/',
'./datasets/crack/panoptic_train2022/',{})

However I got the following error:

Output exceeds the [size limit](command:workbench.action.openSettings?[). Open the full output data [in a text editor](command:workbench.action.openLargeOutput?9c19538d-6c07-4802-8860-9bb873fd1456)
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
Cell In [9], line 16
     13 register_coco_panoptic("ConcreteCrack",{},"./datasets/crack/train2022/","./datasets/crack/panoptic_train2022/",
     14 "./datasets/crack/annotations/panoptic_train2022.json","./datasets/crack/annotations/instances_train2022.json")
     15 MetadataCatalog.get("ConcreteCrack").thing_classes = ["ConcreteCrack"]
---> 16 dicts = load_coco_panoptic_json('./datasets/crack/annotations/panoptic_train2022.json','./datasets/crack/train2022/',
     17 './datasets/crack/panoptic_train2022/',{})

File ~/vsc/Mask2Former/seg/lib/python3.8/site-packages/detectron2/data/datasets/coco_panoptic.py:51, in load_coco_panoptic_json(json_file, image_dir, gt_dir, meta)
     49     image_file = os.path.join(image_dir, os.path.splitext(ann["file_name"])[0] + ".jpg")
     50     label_file = os.path.join(gt_dir, ann["file_name"])
---> 51     segments_info = [_convert_category_id(x, meta) for x in ann["segments_info"]]
     52     ret.append(
     53         {
     54             "file_name": image_file,
   (...)
     58         }
     59     )
     60 assert len(ret), f"No images found in {image_dir}!"

File ~/vsc/Mask2Former/seg/lib/python3.8/site-packages/detectron2/data/datasets/coco_panoptic.py:51, in <listcomp>(.0)
     49     image_file = os.path.join(image_dir, os.path.splitext(ann["file_name"])[0] + ".jpg")
     50     label_file = os.path.join(gt_dir, ann["file_name"])
---> 51     segments_info = [_convert_category_id(x, meta) for x in ann["segments_info"]]
...
     29             segment_info["category_id"]
     30         ]
     31         segment_info["isthing"] = True

KeyError: 'thing_dataset_id_to_contiguous_id'

How can i fix the problem? thx :)

[Ahmed-G-ElTaher]

@david95j2
How to solve KeyError: 'thing_dataset_id_to_contiguous_id'

Simply try to add classes to metadata brackets it will work

the normal is to write like that :
register_coco_panoptic(
data_name, {} , image_root, panoptic_root, panoptic_json, instances_json=None
)

but to solve it write like this :
register_coco_panoptic(
data_name, {"thing_classes": ["class_1","class_2",…etc.]} , image_root, panoptic_root, panoptic_json, instances_json=None
)