Code for (Visual Boundary Knowledge Translation for Foreground Segmentation, AAAI2021). [https://ojs.aaai.org/index.php/AAAI/article/view/16222].

For packages,

pip install -r requirements.txt

Note that pydensecrf need to be installed from source,

pip install git+https://github.com/lucasb-eyer/pydensecrf.git

Note that full_data.txt maps input image path to ground-truth mask path and we split data into train , val , test in the code. All dataset are organized as follows:

/datasets
    /[HumanMatting]
        /Images
        /Labels
        full_data.txt

Birds: http://www.vision.caltech.edu/visipedia/CUB-200-2011.html

HumanMatting: https://github.com/aisegmentcn/matting_human_datasets

Flowers: http://www.robots.ox.ac.uk/~vgg/data/flowers/

MSRA10K and MSRA-B: https://mmcheng.net/msra10k/

ECSSD and CSSD: http://www.cse.cuhk.edu.hk/leojia/projects/hsaliency/dataset.html

DUT-OMRON:  http://saliencydetection.net/dut-omron/

PASCAL-Context: https://www.cs.stanford.edu/~roozbeh/pascal-context/

HKU-IS: https://i.cs.hku.hk/~gbli/deep_saliency.html

SOD: http://elderlab.yorku.ca/SOD/SOD.zip

THUR15K: https://mmcheng.net/code-data/

SIP1K: https://mmcheng.net/code-data/

CAC: https://github.com/akwasigroch/coattention_object_segmentation

ReDO: https://github.com/mickaelChen/ReDO

SG-One: https://github.com/xiaomengyc/SG-One

PANet: https://github.com/kaixin96/PANet

ALSSS: https://github.com/hfslyc/AdvSemiSeg

USSS: https://github.com/tarun005/USSS_ICCV19

Unet/FPN/LinkNet/PSPNet/PAN: https://github.com/qubvel/segmentation_models.pytorch

DeepLab_V3+: https://github.com/jfzhang95/pytorch-deeplab-xception.git

python main.py --device_id=[gpu_id]

Part of the code is heavily borrowed from the official code release of Deeplab_v3+(https://github.com/jfzhang95/pytorch-deeplab-xception.git)