This is the project page for the paper:
ISTR: End-to-End Instance Segmentation via Transformers,
Jie Hu, Liujuan Cao, Lu Yao, ShengChuan Zhang, Yan Wang, Ke Li, Feiyue Huang, Rongrong Ji, Ling Shao
arXiv 2105.00637
⭐Highlights:
- GPU Friendly: Four 1080Ti/2080Ti GPUs can handle the training for R50, R101 backbones with ISTR.
- High Performance: On COCO test-dev, ISTR-R50-3x gets 46.8/38.6 box/mask AP, and ISTR-R101-3x gets 48.1/39.9 box/mask AP.
- (2021.05.03) The project page for ISTR is avaliable.
| Method | inf. time | box AP | mask AP | download |
|---|---|---|---|---|
| ISTR-R50-3x | 17.8 FPS | 46.8 | 38.6 | model | log |
| ISTR-R101-3x | 13.9 FPS | 48.1 | 39.9 | model | log |
- The inference time is evaluated with a single 2080Ti GPU.
- We use the models pre-trained on ImageNet using torchvision. The ImageNet pre-trained ResNet-101 backbone is obtained from SparseR-CNN.
The codes are built on top of Detectron2, SparseR-CNN, and AdelaiDet.
- Python=3.8
- PyTorch=1.6.0, torchvision=0.7.0, cudatoolkit=10.1
- OpenCV for visualization
- Install the repository (we recommend to use Anaconda for installation.)
conda create -n ISTR python=3.8 -y
conda activate ISTR
conda install pytorch==1.6.0 torchvision==0.7.0 cudatoolkit=10.1 -c pytorch
pip install opencv-python
pip install scipy
pip install shapely
git clone https://github.com/hujiecpp/ISTR.git
cd ISTR
python setup.py build develop
- Link coco dataset path
ln -s /coco_dataset_path/coco ./datasets
- Train ISTR (e.g., with ResNet50 backbone)
python projects/ISTR/train_net.py --num-gpus 4 --config-file projects/ISTR/configs/ISTR-R50-3x.yaml
- Evaluate ISTR (e.g., with ResNet50 backbone)
python projects/ISTR/train_net.py --num-gpus 4 --config-file projects/ISTR/configs/ISTR-R50-3x.yaml --eval-only MODEL.WEIGHTS ./output/model_final.pth
- Visualize the detection and segmentation results (e.g., with ResNet50 backbone)
python demo/demo.py --config-file projects/ISTR/configs/ISTR-R50-3x.yaml --input input1.jpg --output ./output --confidence-threshold 0.4 --opts MODEL.WEIGHTS ./output/model_final.pth
If our paper helps your research, please cite it in your publications:
@article{hu2021ISTR,
title={ISTR: End-to-End Instance Segmentation via Transformers},
author={Hu, Jie and Cao, Liujuan and Yao, Lu and Zhang, ShengChuan and Li, Ke and Huang, Feiyue and Ji, Rongrong and Shao, Ling},
journal={arXiv preprint arXiv:2105.00637},
year={2021}
}