Awesome
SG-Net: Spatial Granularity Network for One-Stage Video Instance Segmentation
<p align="center"> <img src="figs/frame.PNG" width="80%"> </p>Introduction
We approach the [VIS task] from a new perspective and propose a one-stage spatial granularity network (SG-Net) (as shown in the above figure). This repo is the implementation of CVPR 2021 paper "SG-Net: Spatial Granularity Network for One-Stage Video Instance Segmentation." [pdf]
Installation
Please find detailed steps Here for installation and dataset preparation.
Train
Please find details Here for step-by-step instructions.
Inference
Please refer to Here for inference.
License
SG-Net
is released under the MIT license.
Citation
If you find this repo useful for your research, please consider citing the paper
@article{liu2021sg,
title={SG-Net: Spatial Granularity Network for One-Stage Video Instance Segmentation},
author={Liu, Dongfang and Cui, Yiming and Tan, Wenbo and Chen, Yingjie},
journal={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
year={2021}
}
Acknowledgements
We truely thanksful of the following piror efforts in terms of knowledge contributions and open-source repos.
- BlendMask: Top-down meets bottom-up for instance segmentation (CVPR'20) [paper] [official code]
- Video Instance Segmenation (ICCV'19) [paper] [official code]