Skip to content

[IEEE J-STARS 2024] [ADSTNet] Adaptive Dual-Stream Sparse Transformer Network for Salient Object Detection in Optical Remote Sensing Images.

Notifications You must be signed in to change notification settings

JieZzzoo/ADSTNet

Repository files navigation

ADSTNet

This project provides the code and results for 'Adaptive Dual-Stream Sparse Transformer Network for Salient Object Detection in Optical Remote Sensing Images', IEEE J-STARS, 2024 (IEEE Link).

Network Architecture

ADSTNet.png

Datasets

The datasets utilized in this work can be accessed from BaiDuYunlink (code:2r9f), including ORSSD, EORSSD, ORSI-4199 and RSISOD.

Saliency maps

We provide saliency maps of our ADSTNet based on Res2Net in BaiDuYunlink (code:ADST) on ORSSD, EORSSD and ORSI-4199.

Evaluation Tool

You can use the evaluation tool (MATLAB version) to evaluate the above saliency maps.

Table.png

Fig6.png

News 🚩

We provide saliency maps of ADSTNet base on the others backbone (VGG and ResNet) in BaiDuYunlink (code:ADST).

We provide saliency maps of some semantic segmentation methods in BaiDuYunlink (code:ADST).

Citation

If you find this work interesting and use our dataset in your research, please cite:

@article{zhao2024adaptive,
  title={Adaptive Dual-Stream Sparse Transformer Network for Salient Object Detection in Optical Remote Sensing Images},
  author={Zhao, Jie and Jia, Yun and Ma, Lin and Yu, Lidan},
  journal={IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
  volume={17},
  pages={5173--5192},
  year={2024},
  publisher={IEEE}
}

If you encounter any problems with the code, want to report bugs, etc.

Please contact me at [email protected].

About

[IEEE J-STARS 2024] [ADSTNet] Adaptive Dual-Stream Sparse Transformer Network for Salient Object Detection in Optical Remote Sensing Images.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published