OPNet: Deep Occlusion Perception Network with Boundary Awareness for Amodal Instance Segmentation

This is the PyTorch implementation of OPNet built on the open-source detectron2.

"OPNet: Deep Occlusion Perception Network with Boundary Awareness for Amodal Instance Segmentation"

Installation conda create -n opnet python=3.7 -y source activate opnet pip install torch==1.11.0+cu113 torchvision==0.12.0+cu113 -f https://download.pytorch.org/whl/torch_stable.html pip install ninja yacs cython matplotlib tqdm pip install opencv-python==4.4.0.40 pip install scikit-image export INSTALL_DIR=$PWD cd $INSTALL_DIR git clone https://github.com/cocodataset/cocoapi.git cd cocoapi/PythonAPI python setup.py build_ext install cd $INSTALL_DIR git clone https://github.com/ZitengXue/OPNet.git cd OPNet/ pip install -r requirements.txt python3 setup.py build develop Dataset Preparation D2SA dataset The D2S Amodal dataset could be found at mvtec-d2sa. KINS dataset Download the Images from KITTI dataset. The Amodal Annotations could be found at KINS dataset CWALT dataset Follow WALT,we will release soon. Expected folder structure for each dataset $PROJECT_DIR/ |OPNet/ |-detectron2/ |-- data/ |---- datasets/ |------- KINS/ |---------- train_imgs/ |---------- test_imgs/ |---------- annotations/ |------------- train.json |------------- test.json |------- D2SA/ |-------- annotations/ |-------- images/ |------- CWALT/ |... Then, See here for more details on data registration Multi-GPU Training and evaluation on Validation set cd OPNet For example, on the D2SA dataset, we show how to train our method and other baselines. 1.Train our model (ResNet50 backbone) on D2SA dataset : CUDA_VISIBLE_DEVICES=0,1 python tools/train_net.py --config-file configs/D2SA-AmodalSegmentation/bounday.yaml--num-gpus 2 Test If you want to eval your saved checkpoints: python tools/train_net.py --config-file configs/{your_yaml_file} --eval-only MODEL.WEIGHTS {your_OUTPUT_DIR}/model_final.pth' --num-gpus x Acknowledgement Our work benefits a lot from BCNet and VRSP-Net. Thanks for their wonderful works.