- Image panoptic segmentation and video panoptic segmentation
- To run:
- Clone repository.
- Follow environment setup instructions below,
- Run "python main.py" in the terminal!
- Credit: M. Haroon Shakeel (Medium: https://medium.com/@haroonshakeel/detectron2-setup-on-windows-10-and-linux-407e5382df1)
- Instructions: (estimated time to complete: 5 minutes)
- Download and Install Anaconda
- Open Anaconda Prompt and create a new virtual environment by using the command: conda create -n detectron_env python=3.8
- Activate the environment: conda activate detectron_env
- Once done, install cython pip install cython
- Install Pytorch and CUDA: conda install pytorch torchvision torchaudio cudatoolkit=11.0 -c pytorch
- Clone Detectron2 repository: git clone https://github.com/facebookresearch/detectron2.git
- Got to the downloaded repository folder: cd detectron2
- Install dependencies: pip install -e .
- Then install OpenCV pip install opencv-python