FAPIS

The official implementation of the CVPR 2021 paper FAPIS: a Few-shot Anchor-free Part-based Instance Segmenter

drawing

Introduction

This repo is primarily based on the Pytorch implementation of Siamese Mask-RCNN and we use mmdetection toolbox to finish it.

The official code of Siamese Mask-RCNN can be found in siamese mask-rcnn

Installation

Please follow the installation in README_mmdetection.md, to compile the necessary libraries, please read the compile.sh file

Prepare COCO dataset

ln -s $path/to/coco data/coco

Training

Please read train_FAPISv2.sh for some sample commands

Testing

Please read test_FAPISv2.sh for some sample commands

and run run.sh the results will be saved in results.txt file

Visualize the results

python tools/test.py configs/FAPISv2_fcos_r50_caffe_fpn_gn_1x_4gpu.py work_dirs/FAPISv2_fcos_use_rf_mask_constrain_parts_unet_dist_part_0/latest.pth --show

Citation

Our paper:

@inproceedings{nguyen2021fapis,
  title={FAPIS: A Few-shot Anchor-free Part-based Instance Segmenter},
  author={Nguyen, Khoi and Todorovic, Sinisa},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={11099--11108},
  year={2021}
}

The Siamese Mask-RCNN paper:

@article{michaelis_one-shot_2018,
    title = {One-Shot Instance Segmentation},
    author = {Michaelis, Claudio and Ustyuzhaninov, Ivan and Bethge, Matthias and Ecker, Alexander S.},
    year = {2018},
    journal = {arXiv},
    url = {http://arxiv.org/abs/1811.11507}
}

This project is based on mmdetection toolbox.

@article{mmdetection,
  title   = {{MMDetection}: Open MMLab Detection Toolbox and Benchmark},
  author  = {Kai Chen, Jiaqi Wang, Jiangmiao Pang, Yuhang Cao, Yu Xiong, Xiaoxiao Li,
             Shuyang Sun, Wansen Feng, Ziwei Liu, Jiarui Xu, Zheng Zhang, Dazhi Cheng,
             Chenchen Zhu, Tianheng Cheng, Qijie Zhao, Buyu Li, Xin Lu, Rui Zhu, Yue Wu,
             Jifeng Dai, Jingdong Wang, Jianping Shi, Wanli Ouyang, Chen Change Loy, Dahua Lin},
  journal = {arXiv preprint arXiv:1906.07155},
  year    = {2019}
}

Thanks for their contributions

GitHub

View Github