Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/segmentationblwx/cssegmentation
CSSegmentation: An Open Source Continual Semantic Segmentation Toolbox Based on PyTorch.
https://github.com/segmentationblwx/cssegmentation
benchmark continual-learning pytorch semantic-segmentation
Last synced: 3 days ago
JSON representation
CSSegmentation: An Open Source Continual Semantic Segmentation Toolbox Based on PyTorch.
- Host: GitHub
- URL: https://github.com/segmentationblwx/cssegmentation
- Owner: SegmentationBLWX
- License: apache-2.0
- Created: 2023-03-19T07:14:39.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-02-06T16:10:33.000Z (8 months ago)
- Last Synced: 2024-09-30T16:37:41.870Z (3 days ago)
- Topics: benchmark, continual-learning, pytorch, semantic-segmentation
- Language: Python
- Homepage: https://cssegmentation.readthedocs.io/en/latest/
- Size: 376 KB
- Stars: 27
- Watchers: 1
- Forks: 4
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- Funding: .github/FUNDING.yml
- License: LICENSE
Awesome Lists containing this project
README
[](https://cssegmentation.readthedocs.io/en/latest/)
[](https://pypi.org/project/cssegmentation/)
[](https://pypi.org/project/cssegmentation)
[](https://github.com/SegmentationBLWX/cssegmentation/blob/master/LICENSE)
[](https://pypi.org/project/cssegmentation/)
[](https://pypi.org/project/cssegmentation/)
[](https://github.com/SegmentationBLWX/cssegmentation/issues)
[](https://github.com/SegmentationBLWX/cssegmentation/issues)
Documents: https://cssegmentation.readthedocs.io/en/latest/
## Introduction
CSSegmentation: An Open Source Continual Semantic Segmentation Toolbox Based on PyTorch.
You can star this repository to keep track of the project if it's helpful for you, thank you for your support.
## Major Features
- **High Performance**
The performance of re-implemented CSS algorithms is better than or comparable to the original paper.
- **Modular Design and Unified Benchmark**
Various CSS methods are unified into several specific modules.
Benefiting from this design, CSSegmentation can integrate a great deal of popular and contemporary continual semantic segmentation frameworks and then, train and test them on unified benchmarks.
- **Fewer Dependencies**
CSSegmentation tries its best to avoid introducing more dependencies when reproducing novel continual semantic segmentation approaches.
## Benchmark and Model Zoo
#### Supported Encoder
| Encoder | Model Zoo | Paper Link | Code Snippet |
| :-: | :-: | :-: | :-: |
| ResNet | [click](./docs/modelzoo/mib) | [CVPR 2016](https://arxiv.org/pdf/1512.03385.pdf) | [click](./csseg/modules/models/encoders/resnet.py) |
#### Supported Decoder
| Decoder | Model Zoo | Paper Link | Code Snippet |
| :-: | :-: | :-: | :-: |
| Deeplabv3 | [click](./docs/modelzoo/mib) | [ArXiv 2017](https://arxiv.org/pdf/1706.05587.pdf) | [click](./csseg/modules/models/decoders/aspphead.py) |
#### Supported Runner
| Runner | Model Zoo | Paper Link | Code Snippet |
| :-: | :-: | :-: | :-: |
| EWF | [click (under developing)](./docs/modelzoo/ewf) | [CVPR 2023](https://openaccess.thecvf.com/content/CVPR2023/papers/Xiao_Endpoints_Weight_Fusion_for_Class_Incremental_Semantic_Segmentation_CVPR_2023_paper.pdf) | [click](./csseg/modules/runners/ewf.py) |
| UCD | [click (under developing)](./docs/modelzoo/ucd) | [TPAMI 2022](https://arxiv.org/pdf/2203.14098.pdf) | [click](./csseg/modules/runners/ucd.py) |
| RCIL | [click (under developing)](./docs/modelzoo/rcil) | [CVPR 2022](https://arxiv.org/pdf/2203.05402.pdf) | [click](./csseg/modules/runners/rcil.py) |
| REMINDER | [click (under developing)](./docs/modelzoo/reminder) | [CVPR 2022](https://openaccess.thecvf.com/content/CVPR2022/papers/Phan_Class_Similarity_Weighted_Knowledge_Distillation_for_Continual_Semantic_Segmentation_CVPR_2022_paper.pdf) | [click](./csseg/modules/runners/reminder.py) |
| CAF | [click (under developing)](./docs/modelzoo/caf) | [TMM 2022](https://arxiv.org/pdf/2202.00432.pdf) | [click](./csseg/modules/runners/caf.py) |
| SDR | [click (under developing)](./docs/modelzoo/sdr) | [CVPR 2021](https://arxiv.org/pdf/2103.06342.pdf) | [click](./csseg/modules/runners/sdr.py) |
| PLOP | [click](./docs/modelzoo/plop) | [CVPR 2021](https://arxiv.org/pdf/2011.11390.pdf) | [click](./csseg/modules/runners/plop.py) |
| MIB | [click](./docs/modelzoo/mib) | [CVPR 2020](https://arxiv.org/pdf/2002.00718.pdf) | [click](./csseg/modules/runners/mib.py) |
| ILT | [click](./docs/modelzoo/ilt) | [ICCVW 2019](https://arxiv.org/pdf/1907.13372.pdf) | [click](./csseg/modules/runners/ilt.py) |
#### Supported Datasets
| Dataset | Project Link | Paper Link | Code Snippet |
| :-: | :-: | :-: | :-: |
| ADE20k | [Click](https://groups.csail.mit.edu/vision/datasets/ADE20K/) | [CVPR 2017](https://arxiv.org/pdf/1608.05442.pdf) | [Click](./csseg/modules/datasets/ade20k.py) |
| PASCAL VOC | [Click](http://host.robots.ox.ac.uk/pascal/VOC/) | [IJCV 2010](http://host.robots.ox.ac.uk/pascal/VOC/pubs/everingham10.pdf) | [Click](./csseg/modules/datasets/voc.py) |
## Citation
If you use this framework in your research, please cite this project:
```
@misc{csseg2023,
author = {Zhenchao Jin},
title = {CSSegmentation: An Open Source Continual Semantic Segmentation Toolbox Based on PyTorch},
year = {2023},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/SegmentationBLWX/cssegmentation}},
}
```