This is a PyTorch implementation of Deep GrabCut, for object segmentation. We use DeepLab-v2 instead of DeconvNet in this repository.

The code was tested with Python 3.5. To use this code, please do:

0. Clone the repo:

   git clone https://github.com/jfzhang95/DeepGrabCut-PyTorch
   cd DeepGrabCut-PyTorch

1. Install dependencies:

   pip install pytorch torchvision -c pytorch
   pip install matplotlib opencv pillow

2. You can download pretrained model from GoogleDrive, and then put the model into models/

3. To try the demo of Deep GrabCut, please run:

   python demo.py

If installed correctly, the result should look like this:

To train Deep GrabCut on VOC (or VOC + SBD), please follow these additional steps:

4. Download the pre-trained PSPNet model for semantic segmentation, taken from this repository.

   cd models/
   chmod +x download_pretrained_psp_model.sh
   ./download_pretrained_psp_model.sh
   cd ..

5. Set the paths in mypath.py, so that they point to the location of VOC/SBD dataset.

6. Run python train.py to train Deep Grabcut.

7. If you want to train model on COCO dataset, you should first config COCO dataset path in mypath.py, and then run python train_coco.py to train model.