[Mask DINO] Environment Setting & Dataloader with Custom Dataset

https://github.com/IDEA-Research/MaskDINO

GitHub - IDEA-Research/MaskDINO: [CVPR 2023] Official implementation of the paper "Mask DINO: Towards A Unified Transformer-base

[Environment Info]

Anaconda Environment

RTX4090 => python=3.8, pytorch=1.9.0, cuda=11.8

 

 

[Error]

#######################################################################################

"ImportError: cannot import name 'cached_property' from 'functools' (/opt/conda/lib/python3.7/functools.py)"

 

python 3.8 이상의 버전 필요

#######################################################################################

 

[Environment Setting]

pip install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html

python -m pip install detectron2 -f \
  https://dl.fbaipublicfiles.com/detectron2/wheels/cu111/torch1.9/index.html
  
git clone https://github.com/facebookresearch/detectron2.git
cd detectron2
pip install -e .
pip install git+https://github.com/cocodataset/panopticapi.git
pip install git+https://github.com/mcordts/cityscapesScripts.git

git clone https://github.com/IDEA-Research/MaskDINO.git
cd MaskDINO
pip install -r requirements.txt
cd maskdino/modeling/pixel_decoder/ops
sh make.sh

 

 

[Dataloader]

train_net.py 안에 제일 상위에 dataset register 해줘야 함

ex)

register_coco_instances("my_dataset", {}, "/home/.../train.json", "/home/.../train/images/")
MetadataCatalog.get("my_dataset").thing_classes = ["dog"]


if__name__=="__main__":

    ...

    ...

    ...