Chaewon Lee, Seon-Ho Lee, and Chang-Su Kim

Official code for "MFP: Making Full Use of Probability Maps for Interactive Image Segmentation"[paper]

• PyTorch 1.11.0
• CUDA 11.3
• CuDNN 8.2.0
• python 3.8

Create conda environment:

    $ conda create -n MFP python=3.8 anaconda
    $ conda activate MFP
    $ conda install pytorch==1.11.0 torchvision==0.12.0 cudatoolkit=11.3 -c pytorch
    $ pip install -r requirements.txt

Download repository:

    $ git clone https://github.com/cwlee00/MFP.git

Download weights:

MFP model Google Drive

For evaluation, please download the datasets and models, and then configure the path in config.yml

python scripts/evaluate_model.py NoBRS \
--gpu=0 \
--checkpoint=./weights/mfp_models/MFP_vit_base(cocolvis).pth \
--eval-mode=cvpr \
--datasets=GrabCut,Berkeley,DAVIS,SBD

For training, please download the MAE pretrained weights (click to download: ViT-Base) and configure the dowloaded path in config.yml.

python train.py models/iter_mask/plainvit_base448_cocolvis_itermask_prevMod.py \
--batch-size=8 \
--ngpus=1

Please cite the following paper if you feel this repository useful.

    @InProceedings{Lee_2024_CVPR,
    author    = {Lee, Chaewon and Lee, Seon-Ho and Kim, Chang-Su},
    title     = {MFP: Making Full Use of Probability Maps for Interactive Image Segmentation},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2024},
    pages     = {4051-4059}
    }

Our project is developed based on RITM and SimpleClick. We would like to show sincere thanks to the contributors.