DARNet-CD: A Densely Attentive Refinement Network for Change Detection based on Very-High-Resolution Bi-Temporal Remote Sensing Images

This repo is the official implementation for DARNet proposed in the journal article "A Densely Attentive Refinement Network for Change Detection based on Very-High-Resolution Bi-Temporal Remote Sensing Images" accepted by IEEE Transactions on Geoscience and Remote sensing. More details about this work are described in the paper (https://ieeexplore.ieee.org/document/9734050).

Requirements

• Python 3.7
• PyTorch 1.7.0
• Torchvision 0.8.0

Datasets

CDD

Change detection in remote sensing images using conditional adversarial networks

LEVIR-CD

A Spatial-Temporal Attention-Based Method and a New Dataset for Remote Sensing Image Change Detection

SYSUCD

A Deeply Supervised Attention Metric-Based Network and an Open Aerial Image Dataset for Remote Sensing Change Detection

Basic Usage

Prepare the training/validation/testing datasets as the examples in data directory.

Train

python train.py

Evaluate

python evaluate.py

Inference

python inference.py

Pretrained weights

The pretrained models can be downloaded soon.

Citation

If you find this code useful and utilize it in your own research, please consider citing our article with the following bibtex:

@ARTICLE{9734050,
  author={Li, Ziming and Yan, Chenxi and Sun, Ying and Xin, Qinchuan},
  journal={IEEE Transactions on Geoscience and Remote Sensing}, 
  title={A Densely Attentive Refinement Network for Change Detection based on Very-High-Resolution Bi-Temporal Remote Sensing Images}, 
  year={2022},
  volume={},
  number={},
  pages={1-1},
  doi={10.1109/TGRS.2022.3159544}}

Contact

If you have any question about this code, please contact Ziming Li: lizm9@mail2.sysu.edu.cn