installation.md

Installation

The project is based on the pytorch 1.8.1 with python 3.7.

1. Clone the Git repo

$ git clone https://github.com/yyliu01/PS-MT.git
$ cd PS-MT

2. Install dependencies

1. create conda env

   $ conda env create -f ps-mt.yml

2. install the torch 1.8.1

   $ conda activate ps-mt
   # IF cuda version < 11.0
   $ pip install torch==1.8.1+cu102 torchvision==0.9.1+cu102 torchaudio==0.8.1 -f https://download.pytorch.org/whl/torch_stable.html
   # IF cuda version >= 11.0 (e.g., 30x or above)
   $ pip install torch==1.8.1+cu111 torchvision==0.9.1+cu111 torchaudio==0.8.1 -f https://download.pytorch.org/whl/torch_stable.html

3. Prepare dataset

1. please download the VOC12 and Cityscapes dataset (gt_Fine).
2. organize the filepath to point out the root folder for each of them, in the config file.

    "train_supervised": {
        "data_dir": "path_to/VOCtrainval_11-May-2012",
        ...
    }
    "train_unsupervised": {
        "data_dir": "path_to/VOCtrainval_11-May-2012",
        ...
    }
    "val": {
        "data_dir": "path_to/VOCtrainval_11-May-2012",
        ...
    }

3. (optional) you might need to preprocess Cityscapes labels in here, as we follow the common setting with 19 classes.

please note, our training details (e.g., training id, input resolution, and batch size) follow the previous works CPS and PseudoSeg.