https://github.com/henghuiding/bfp
[ICCV2019] Boundary-Aware Feature Propagation for Scene Segmentation
https://github.com/henghuiding/bfp
bfp iccv2019 pytorch semantic-segmentation
Last synced: 6 months ago
JSON representation
[ICCV2019] Boundary-Aware Feature Propagation for Scene Segmentation
- Host: GitHub
- URL: https://github.com/henghuiding/bfp
- Owner: henghuiding
- License: mit
- Created: 2020-04-21T16:29:15.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2020-04-21T16:31:52.000Z (over 5 years ago)
- Last Synced: 2025-03-25T10:51:10.462Z (7 months ago)
- Topics: bfp, iccv2019, pytorch, semantic-segmentation
- Language: Python
- Homepage:
- Size: 245 KB
- Stars: 80
- Watchers: 2
- Forks: 9
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# [Boundary-Aware Feature Propagation for Scene Segmentation](http://openaccess.thecvf.com/content_ICCV_2019/papers/Ding_Boundary-Aware_Feature_Propagation_for_Scene_Segmentation_ICCV_2019_paper.pdf)
1. Install pytorch
- The code is implemented on python3.6 and official [Pytorch](https://github.com/pytorch/pytorch/tree/fd25a2a86c6afa93c7062781d013ad5f41e0504b#from-source).
- Please install [PyTorch-Encoding](https://github.com/zhanghang1989/PyTorch-Encoding) for Synchronized BN.
2. Usage
- Training:
```shell
python train.py --model bfp --dataset pascalcontext --checkname bfp101 --batch-size 12
```
- Testing:
```shell
python test.py --model bfp --dataset pascalcontext --resume-dir model_path --eval
```
3. Citations
- Please consider citing our paper in your publications if the project helps your research.
```
@inproceedings{ding2019bfp,
title={Boundary-aware feature propagation for scene segmentation},
author={Ding, Henghui and Jiang, Xudong and Liu, Ai Qun and Thalmann, Nadia Magnenat and Wang, Gang},
booktitle={Proceedings of the IEEE International Conference on Computer Vision},
pages={6819--6829},
year={2019}
}
```