Discrepency in the expected range of category_ids

[skhushu]

My team and I are running a training on custom dataset.

We realized there is a discrepency in detectron2 code and documentation on whether the category_ids should be in the range [0, num_categories-1] or [1, num_categories]

The documentation states that the category_ids should be in the range [0, num_categories-1]

Lines 82-98 in detectron2/data/datasets/coco.py
The description says that category_ids are expected to be in [0,80) but the code throws a warning if it's not in [1, 80]. Funny enough it re-maps it to [0, 80) after that warning.

        # In COCO, certain category ids are artificially removed,
        # and by convention they are always ignored.
        # We deal with COCO's id issue and translate
        # the category ids to contiguous ids in [0, 80).

        # It works by looking at the "categories" field in the json, therefore
        # if users' own json also have incontiguous ids, we'll
        # apply this mapping as well but print a warning.
        if not (min(cat_ids) == 1 and max(cat_ids) == len(cat_ids)):
            if "coco" not in dataset_name:
                logger.warning(
                    """
Category ids in annotations are not in [1, #categories]! We'll apply a mapping for you.
"""
          )
        id_map = {v: i for i, v in enumerate(cat_ids)}
        meta.thing_dataset_id_to_contiguous_id = id_map

I think for consistency, the code checking for min and max should be changed to if not (min(cat_ids) == 0 and max(cat_ids) == len(cat_ids)-1)

Slightly unrelated concern is that in coco_evaluation, the num_classes are being read from the dataset and not the config which can be an issue if the dataset doesn't have all classes in it.
Lines 230-235 in evaluation/coco_evaluation.py

        # unmap the category ids for COCO
        if hasattr(self._metadata, "thing_dataset_id_to_contiguous_id"):
            dataset_id_to_contiguous_id = self._metadata.thing_dataset_id_to_contiguous_id
            all_contiguous_ids = list(dataset_id_to_contiguous_id.values())
            num_classes = len(all_contiguous_ids)
            assert min(all_contiguous_ids) == 0 and max(all_contiguous_ids) == num_classes - 1

Environment:

sys.platform                     linux
Python                           3.8.10 (default, May 26 2023, 14:05:08) [GCC 9.4.0]
numpy                            1.24.2
detectron2                       0.6 @/home/ubuntu/detectron2/detectron2
Compiler                         GCC 9.4
CUDA compiler                    CUDA 11.7
detectron2 arch flags            7.5
DETECTRON2_ENV_MODULE            <not set>
PyTorch                          1.13.1+cu117 @/home/ubuntu/venv/lib/python3.8/site-packages/torch
PyTorch debug build              False
torch._C._GLIBCXX_USE_CXX11_ABI  False
GPU available                    Yes
GPU 0                            Tesla T4 (arch=7.5)
Driver version                   515.65.01
CUDA_HOME                        /usr/local/cuda
Pillow                           9.4.0
torchvision                      0.14.1+cu117 @/home/ubuntu/venv/lib/python3.8/site-packages/torchvision
torchvision arch flags           3.5, 5.0, 6.0, 7.0, 7.5, 8.0, 8.6
fvcore                           0.1.5.post20221221
iopath                           0.1.9
cv2                              4.5.5
-------------------------------  ------------------------------------------------------------------------------------------------
PyTorch built with:
  - GCC 9.3
  - C++ Version: 201402
  - Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v2.6.0 (Git Hash 52b5f107dd9cf10910aaa19cb47f3abf9b349815)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - LAPACK is enabled (usually provided by MKL)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - CUDA Runtime 11.7
  - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86
  - CuDNN 8.6  (built against CUDA 11.8)
    - Built with CuDNN 8.5
  - Magma 2.6.1
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.7, CUDNN_VERSION=8.5.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.13.1, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, 

Instructions To Reproduce the Issue:

1. Set the 'num_categories` in a detection problem's config file to 80.
2. Create coco instances file with category_ids starting from 0

[github-actions]

You've chosen to report an unexpected problem or bug. Unless you already know the root cause of it, please include details about it by filling the issue template.
The following information is missing: "Instructions To Reproduce the Issue and Full Logs"; "Your Environment";