Introduction

sssegmentation is a general framework for our research on strongly supervised semantic segmentation.

Documents

In English

https://github.com/SegmentationBLWX/sssegmentation/tree/main/docs

Supported

Supported Backbones

• HRNet
• ResNet
• ResNeSt
• MobileNetV2
• MobileNetV3

Supported Models

• FCN
• CE2P
• CCNet
• DANet
• GCNet
• DMNet
• EncNet
• OCRNet
• DNLNet
• ANNNet
• EMANet
• PSPNet
• PSANet
• APCNet
• UPerNet
• Deeplabv3
• SemanticFPN
• NonLocalNet
• Deeplabv3Plus

Supported Datasets

• LIP
• ATR
• CIHP
• ADE20k
• MS COCO
• MHPv1&v2
• CityScapes
• Supervisely
• SBUShadow
• PASCAL VOC
• COCOStuff10k
• Pascal Context

Citation

If you use this framework in your research, please cite this project.

@misc{ssseg2020,
    author = {Zhenchao Jin},
    title = {SSSegmentation: A general framework for strongly supervised semantic segmentation},
    year = {2020},
    publisher = {GitHub},
    journal = {GitHub repository},
    howpublished = {\url{https://github.com/SegmentationBLWX/sssegmentation}},
}

References

[1]. https://github.com/open-mmlab/mmcv
[2]. https://github.com/open-mmlab/mmsegmentation