# SAM-CD Pytorch codes of **Adapting Segment Anything Model for Change Detection in HR Remote Sensing Images** [[paper](https://ieeexplore.ieee.org/document/10443350)] ![alt text](https://github.com/ggsDing/SAM-CD/blob/main/flowchart.png) The SAM-CD adopts [FastSAM](https://github.com/CASIA-IVA-Lab/FastSAM) as the visual encoder with some modifications. ## 2024-4-30 Update: SAM-CD now supports access to [efficientSAM](https://github.com/yformer/EfficientSAM). Check the updated model at ```models/effSAM_CD.py``` (prior installation of efficientSAM at the project folder is required). However, direct integration of efficientSAM may cause an accuracy drop, so there is space to further improve the SAM-CD architecture. ## How to Use 1. Installation * Install [FastSAM](https://github.com/CASIA-IVA-Lab/FastSAM) following the instructions. * Modify the Ultralytics source files following the instructions at: ['SAM-CD/models/FastSAM/README.md'](https://github.com/ggsDing/SAM-CD/blob/main/models/FastSAM/README.md). 2. Dataset preparation. * Please split the data into training, validation and test sets and organize them as follows: ``` YOUR_DATA_DIR ├── ... ├── train │ ├── A │ ├── B │ ├── label ├── val │ ├── A │ ├── B │ ├── label ├── test │ ├── A │ ├── B │ ├── label ``` * Find change line 13 in [SAM-CD/datasets/Levir_CD.py](https://github.com/ggsDing/SAM-CD/blob/main/datasets/Levir_CD.py) (or other data-loading .py files), change `/YOUR_DATA_ROOT/` to your local dataset directory. 3. Training classic CD training: `python train_CD.py` training CD with the proposed task-agnostic semantic learning: `python train_SAM_CD.py` line 16-45 are the major training args, which can be changed to load different datasets, models and adjust the training settings. 5. Inference and evaluation inference on test sets: set the chkpt_path and run `python pred_CD.py` evaluation of accuracy: set the prediction dir and GT dir, and run `python eval_CD.py` (More details to be added...) ## Dataset Download In the following, we summarize links to some frequently used CD datasets: * [LEVIR-CD](https://justchenhao.github.io/LEVIR/) * [WHU-CD](https://study.rsgis.whu.edu.cn/pages/download/) [(baidu)](https://pan.baidu.com/s/1A0_xbV4ZktWCbL3j94CInA?pwd=WHCD ) * [CLCD (Baidu)](https://pan.baidu.com/s/1iZtAq-2_vdqoz1RnRtivng?pwd=CLCD) * [S2Looking](https://github.com/S2Looking/Dataset) * [SYSU-CD](https://github.com/liumency/SYSU-CD) ## Pretrained Models For readers to easily evaluate the accuracy, we provide the trained weights of the SAM-CD. [Drive](https://drive.google.com/drive/folders/14tNtID43o-LHs8VaMK5jai1Uf8NqMDAW?usp=sharing) [Baidu](https://pan.baidu.com/s/1V25TFGL5V05ZB5ttFXFSEA?pwd=SMCD) (pswd: SMCD) ## Cite SAM-CD If you find this work useful or interesting, please consider citing the following BibTeX entry. ``` @article{ding2024adapting, title={Adapting Segment Anything Model for Change Detection in HR Remote Sensing Images}, author={Ding, Lei and Zhu, Kun and Peng, Daifeng and Tang, Hao and Yang, Kuiwu and Bruzzone, Lorenzo}, journal={IEEE Transactions on Geoscience and Remote Sensing}, year={2024}, volume={62}, pages={1-11}, doi={10.1109/TGRS.2024.3368168} } ```