C2F-SemiCD-and-C2FNet:https://chengxihan.github.io/

C2F-SemiCD：A Semi-Supervised CD method

C2FNet：A Supervised CD method

The Pytorch implementation for: “C2F-SemiCD and C2FNet: A coarse-to-fine semi-supervised change detection method based on consistency regularization in High-Resolution Remote-Sensing Images,” IEEE Transactions on Geoscience and Remote Sensing（TGRS）, 2024, DOI: 10.1109/TGRS.2024.3370568.Chengxi Han,Chen Wu,Meiqi Hu,Jiepan Li,Hongruixuan Chen

Requirement

-Pytorch 1.8.0  
-torchvision 0.9.0  
-python 3.8  
-opencv-python  4.5.3.56  
-tensorboardx 2.4  
-Cuda 11.3.1  
-Cudnn 11.3  

Training, Test and Visualization Process

1.Semi-supervised training:
python train_C2F-SemiCD.py --epoch 2 --batchsize 16 --gpu_id '1' --data_name 'WHU' --train_ratio 0.05 --model_name 'SemiModel_noema04'

2.Fully supervised training:
python train_C2FNet.py --epoch 2 --batchsize 16 --gpu_id '1' --data_name 'WHU' --train_ratio 0.05 --model_name 'SemiModel_noema04'

3.Ablation experiment training:
python train_C2F-SemiCD_Ablation.py --epoch 2 --batchsize 16 --gpu_id '1' --data_name 'WHU' --train_ratio 0.05 --model_name 'SemiModel_noema04'

1.Semi-supervised testing:
python test_C2F-SemiCD.py --gpu_id '2' --data_name 'WHU' --model_name 'SemiModel_noema04'

2.Fully supervised testing:
python test_C2FNet.py --gpu_id '2' --data_name 'WHU' --model_name 'SemiModel_noema04'

3.Ablation experiments test:
python test_Ablation.py --gpu_id '2' --data_name 'WHU' --model_name 'SemiModel_noema04'

4.Feature visualization:
python test_visualisation.py --gpu_id '2' --data_name 'WHU' --model_name 'SemiModel_noema04'

Test our trained model result

You can directly test our model by our provided training weights in out/. And make sure the weight name is right. Of course, for different methods and datasets, the Dataset mean and std setting is different.

path = opt.weight_dir+'final_epoch99.pt'
parser.add_argument('--save_path', type=str, default='./output/C2F-SemiCD/WHU-5/')  # Semi-supervised models save paths！！
parser.add_argument('--save_path', type=str, default='./output/C2FNet/WHU-5/')  # Fully supervised models save paths！！

Dataset Download

LEVIR-CD：https://justchenhao.github.io/LEVIR/

WHU-CD：http://gpcv.whu.edu.cn/data/building_dataset.html ,our paper split in Baidu Disk,pwd:6969

SYSU-CD: Our paper split in Baidu Disk,pwd:2023

S2Looking-CD: Our paper split in Baidu Disk,pwd:2023

CDD-CD: Our split in Baidu Disk,pwd:2023

DSIFN-CD: Our split in Baidu Disk,pwd:2023

Note: Please crop the LEVIR dataset to a slice of 256×256 before training with it.

And also we provide all test results of our C2F-SemiCD and C2FNet in the output!!!! Download in output or Baidu Disk,pwd:2023 😋😋😋

Dataset Path Setting

 LEVIR-CD or WHU-CD  or GoogleGZ-CD
     |—train  
          |   |—A  
          |   |—B  
          |   |—label  
     |—val  
          |   |—A  
          |   |—B  
          |   |—label  
     |—test  
          |   |—A  
          |   |—B  
          |   |—label

Where A contains images of the first temporal image, B contains images of the second temporal images, and the label contains ground truth maps.

Quantization accuracy

Citation

If you use this code for your research, please cite our papers.

@ARTICLE{10445496,
  author={Han, Chengxi and Wu, Chen and Hu, Meiqi and Li, Jiepan and Chen, Hongruixuan},
  journal={IEEE Transactions on Geoscience and Remote Sensing}, 
  title={C2F-SemiCD: A Coarse-to-Fine Semi-Supervised Change Detection Method Based on Consistency Regularization in High-Resolution Remote Sensing Images}, 
  year={2024},
  volume={62},
  number={},
  pages={1-21},
  keywords={Feature extraction;Training;Remote sensing;Decoding;Semisupervised learning;Data models;Predictive models;Attention mechanism;deep learning;high-resolution remote sensing images;semi-supervised change detection (CD)},
  doi={10.1109/TGRS.2024.3370568}}

Acknowledgments

Thanks for my co-authors Jiepan Li,Haonan Guo, Thanks for their great work!!

Reference

[1] Han, C., Wu, C., Hu, M., Li, J. and Chen, H., 2024. “C2F-SemiCD and C2FNet: A Coarse-to-Fine Semi-Supervised Change Detection Method Based on Consistency Regularization in High-Resolution Remote-Sensing Images,” . IEEE Transactions on Geoscience and Remote Sensing.

[2] C. HAN, C. WU, H. GUO, M. HU, AND H. CHEN, “HANet: A hierarchical attention network for change detection with bi-temporal very-high-resolution remote sensing images,” IEEE J. SEL. TOP. APPL.EARTH OBS. REMOTE SENS., PP. 1–17, 2023, DOI: 10.1109/JSTARS.2023.3264802.

[3] Han, C., Wu, C., Guo, H., Hu, M., Li, J. and Chen, H., 2023. “CGNet: Change guiding network: Incorporating change prior to guide change detection in remote sensing imagery,” . IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[4]Han, C., Wu, C. and Du, B., 2023, HCGMNET: A Hierarchical Change Guiding Map Network For Change Detection, July. In IGARSS 2023-2023 IEEE International Geoscience and Remote Sensing Symposium (pp. 5511-5514). IEEE.

(Don't hesitate to tell me about the latest progress and useful methods in the CD field, I will spare no effort to thank you. Good luck to you guys, I wish we can be the best friend in the CD field.)