Semantic Segmentation for Remote Sensing

This repo is the PyTorch implementation of some works related to remote sensing tasks.

Methods

SAM-based:

MANet (arxiv)

SAM_RS (IEEE TGRS 2024).

Mamba-based:

RS^3Mamba (IEEE GRSL 2024).

Unsupervised Domain Adaptation (UDA):

GLGAN (IEEE TGRS 2024).

MBATA_GAN (IEEE TGRS 2023).

Multimodal Fusion:

MANet (arxiv)

ASMFNet (IEEE JSTARS 2024)

FTransUNet (IEEE TGRS 2024).

CMFNet (IEEE JSTARS 2022).

MSFNet (IEEE IGARSS 2022).

Reference

For UDA:

• TransGAN
• Advent
• AdaSegNet

For Segment Anything Model:

• Medical-SAM-Adapter
• SAM

For Semantic Segmentation:

• TransUNet
• UNetformer
• UCTransNet
• Swin-Transformer
• Swin-Unet
• v-FuseNet
• FuseNet

For Semantic Segmentation based on Mamba:

• VMamba
• VM-UNet
• Swin-UMamba

Datasets

All datasets including ISPRS Potsdam, ISPRS Vaihingen, loveDA can be downloaded here.

Scripts

There are some scripts in utils folder including: split images, merge images, draw loss curves.

Other Works

There are some other works in our group:

Scene Classification: DF4LCZ (IEEE TGRS 2024).

Change Detection: GCD-DDPM (IEEE TGRS 2024), GVLM (ISPRS 2023).

Super-resolution: ASDDPM (IEEE JSTARS Undergoing Review), GCRDN (IEEE JSTARS 2023).

Citations

If these codes are helpful for your study, please cite:

@article{ma2024frequency,
  author={Ma, Xianping and Zhang, Xiaokang and Ding, Xingchen and Pun, Man-On and Ma, Siwei},
  journal={IEEE Transactions on Geoscience and Remote Sensing}, 
  title={Decomposition-Based Unsupervised Domain Adaptation for Remote Sensing Image Semantic Segmentation}, 
  year={2024},
  volume={62},
  number={},
  pages={1-18}
}

@ARTICLE{ma2024rs3mamba,
  author={Ma, Xianping and Zhang, Xiaokang and Pun, Man-On},
  journal={IEEE Geoscience and Remote Sensing Letters}, 
  title={RS3Mamba: Visual State Space Model for Remote Sensing Image Semantic Segmentation}, 
  year={2024},
  volume={},
  number={},
  pages={1-1}
}

@ARTICLE{ma2024ftransunet,
  author={Ma, Xianping and Zhang, Xiaokang and Pun, Man-On and Liu, Ming},
  journal={IEEE Transactions on Geoscience and Remote Sensing}, 
  title={A Multilevel Multimodal Fusion Transformer for Remote Sensing Semantic Segmentation}, 
  year={2024},
  volume={62},
  number={},
  pages={1-15},
  doi={10.1109/TGRS.2024.3373033}
}

@article{ma2024sam,
  author={Ma, Xianping and Wu, Qianqian and Zhao, Xingyu and Zhang, Xiaokang and Pun, Man-On and Huang, Bo},
  journal={IEEE Transactions on Geoscience and Remote Sensing}, 
  title={SAM-Assisted Remote Sensing Imagery Semantic Segmentation With Object and Boundary Constraints}, 
  year={2024},
  volume={62},
  pages={1-16},
}

@article{ma2023unsupervised,
  title={Unsupervised domain adaptation augmented by mutually boosted attention for semantic segmentation of vhr remote sensing images},
  author={Ma, Xianping and Zhang, Xiaokang and Wang, Zhiguo and Pun, Man-On},
  journal={IEEE Transactions on Geoscience and Remote Sensing},
  volume={61},
  pages={1--15},
  year={2023},
  publisher={IEEE}
}

@ARTICLE{ma2024asmfnet,
  author={Ma, Xianping and Xu, Xichen and Zhang, Xiaokang and Pun, Man-On},
  journal={IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing}, 
  title={Adjacent-Scale Multimodal Fusion Networks for Semantic Segmentation of Remote Sensing Data}, 
  year={2024},
  volume={},
  number={},
  pages={1-13}
}

@article{ma2022crossmodal,
  title={A crossmodal multiscale fusion network for semantic segmentation of remote sensing data},
  author={Ma, Xianping and Zhang, Xiaokang and Pun, Man-On},
  journal={IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
  volume={15},
  pages={3463--3474},
  year={2022},
  publisher={IEEE}
}

@inproceedings{ma2022msfnet,
  title={MSFNET: Multi-Stage Fusion Network for Semantic Segmentation of Fine-Resolution Remote Sensing Data},
  author={Ma, Xianping and Zhang, Xiaokang and Pun, Man-On and Liu, Ming},
  booktitle={IGARSS 2022-2022 IEEE International Geoscience and Remote Sensing Symposium},
  pages={2833--2836},
  year={2022},
  organization={IEEE}
}

Contact

Xianping Ma ([email protected]), ([email protected])

Star History