🛠️ Requirements

Environment

Linux system, Windows is not tested, depending on whether and can be installed causal-conv1d and mamba-ssm
Python 3.8+, recommended 3.10
PyTorch 2.0 or higher, recommended 2.1.0
CUDA 11.7 or higher, recommended 12.1

Environment Installation

It is recommended to use Miniconda for installation. The following commands will create a virtual environment named stnr and install PyTorch. In the following installation steps, the default installed CUDA version is 12.1. If your CUDA version is not 12.1, please modify it according to the actual situation.

# Create conda environment
conda create -n stnr python=3.8 -y
conda activate stnr

# Install PyTorch
pip install torch==2.1.0 torchvision==0.14.0 torchaudio==0.13.0

# Install dependencies
pip install causal_conv1d mamba_ssm packaging
pip install timm==0.4.12
pip install pytest chardet yacs termcolor
pip install submitit tensorboardX
pip install triton==2.0.0

# Or simply run
pip install -r requirements.txt

📁 Dataset Preparation

We evaluate our method on five remote sensing change detection datasets: WHU-CD, LEVIR-CD, LEVIR-CD+, SYSU-CD, and SVCD.

Dataset	Link
WHU-CD	Download
LEVIR-CD	Download
LEVIR-CD+	Download
SYSU-CD	Download
SVCD	Download

Please organize the datasets as follows:

${DATASET_ROOT} # Dataset root directory, for example: /home/username/data/LEVIR-CD
├── A
│   ├── train_1_1.png
│   ├── train_1_2.png
│   ├──...
│   ├── val_1_1.png
│   ├── val_1_2.png
│   ├──...
│   ├── test_1_1.png
│   ├── test_1_2.png
│   └── ...
├── B
│   ├── train_1_1.png
│   ├── train_1_2.png
│   ├──...
│   ├── val_1_1.png
│   ├── val_1_2.png
│   ├──...
│   ├── test_1_1.png
│   ├── test_1_2.png
│   └── ...
├── label
│   ├── train_1_1.png
│   ├── train_1_2.png
│   ├──...
│   ├── val_1_1.png
│   ├── val_1_2.png
│   ├──...
│   ├── test_1_1.png
│   ├── test_1_2.png
│   └── ...
├── list
│   ├── train.txt
│   ├── val.txt
│   └── test.txt

🔧 Model Training and Testing

All configuration for model training and testing is stored in the local folder config. Below are the example commands to train and test the model on the LEVIR-CD dataset.

Example of Training on LEVIR-CD Dataset

python train_cd.py --config/levir/levir.json

Example of Training on LEVIR-CD Dataset

python test_cd.py --config/levir/levir_test.json

📂 Project Structure

STNR-Det/
├── dataset/          # Dataset loading and preprocessing
├── model/            # Network architecture (STNR-Det)
├── utils/            # Utility functions
├── weight/           # Pretrained weights
├── main.py           # Main entry point
├── requirements.txt  # Dependencies
└── README.md